• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Pioneer RT-707 service & rebuild

I've completed the playback alignment (balance and EQ). I found the image in an old post of mine useful, in which there is a picture showing which pots are for which channel and purpose. In that post and another post the balance/EQ alignment process is outlined. In the old posts I had a crack at the playback balance & EQ adjustments. I had some luck but had some difficulties. I've now replaced the old VRs and redone the head alignment and things have worked out nicely.

One note regarding the azimuth head adjustment. The key, for me at least, was not worrying about getting the phase of L and R perfectly aligned at the beginning when doing the coarse adjustment (at 8 kHz on my test tape) and just focusing on getting the maximum signal from both channels - don't worry for now if the L and R levels are a bit different (if they are very different you may have to look into other head adjustment as per manual, but otherwise you can balance it up in balance/EQ stage later). I watch the two waveforms (L & R) on the scope. As you adjust the azimuth, the relative phase of L and R channels changes (a lot) and you want to find the maximum signal of both channels. What the manual doesn't tell you is how easy it is to have the phase between the signals 360, or 720 degrees off, especially if you just focus on fine adjustment with 16kHz test tone. Maybe not a big deal since everything is in phase, but anyway, the point is to get the heads aligned so I wanted to do my best. Anyway, so with 8kHz is it a bit easier to do the coarse adjustment, making sure you are near 0 degrees phase difference (and not 360, 720 or so off) and maximum signal, and then at 16kHz you can do the fine adjustment to get as close to 0 as and maximum as possible, once you have got the 'correct peaks' approximately lined up on your scope. You can of course use a relative phase 'x-y' function on your scope if you have it, but I guess I would just say take care you are not an integer*2*pi out of phase. Maybe silly ramblings to the experts here, but useful notes for a newbie like me I think.

Also, some steps of the balance (gain) and EQ setup are tricky since they affect each other, and it is easy to get obsessed with getting it perfect. Don't worry about getting all frequencies perfect at first. You just have to focus on getting the high frequency response point (10kHz) close to the reference point (500 or 1kHz in my case). Focusing on those points got me to the aligned setup quickly with just a few tweaks to get it straight out to 20kHz; whereas worrying about all frequencies at the beginning and trying to get it perfect is the path to madness - it all worked out nicely when i started just focusing on the two main reference points and then tweaked after. Also I noted that on my machine when I balanced gain at 0dB, it L & R would be a bit off at -10dB! It was driving me mad trying to get the balance right at both volume levels, but then at some point I realized it was well within spec and decided get on with my life!

The following graphs tell the story of before and after playback balance/EQ setup. The graphs explain perfectly what I heard in listening tests beforehand (forward directly image shifted to one side, reverse image centered and brighter tone). Note that voltage values in the graphs have more uncertainty at higher frequency (especially from 10kHz and up).

Here is the before and after or the reverse playback.
WmWHDvx.jpg


And here is the result for the forward playback.
0OgVdD9.jpg


Also, I made notes about how stable the amplitudes of the test tones were on playback, especially at high frequency when things start to jump around a bit. Useful info I think. For example my left channel on my reverse direction is a bit more unstable than other 3 channels; I've done my best with the head alignment and I think it is a limitation of the head wear but it's not bad and I have a spare, almost new, set of heads on standby for the future. The rest of the tape path looks good mechanically. Is it worth new rollers if the old ones 'look okay'?

And, and the final point. No matter what you do, if you get the EQ flat within say 100Hz - 20kHz, you have that increase bass below 100Hz. Will be curious to see if it is the same on my other 707s and thus by design... might explain why the 707s sound so pleasing on the bottom end?

Great post! This will be very helpful when I start lining mine up after the Head Block returns from re-lapping. The bass boost is interesting. I don't know if you have that data in terms of dBV or not? Which tape were you using to do the EQ and Balance adjustments? Again, thanks for the through post.
 
Hi,
I tend to prefer the lissajous figure and not looking for peaks. If you do not do a course adjustment at lower frequencies, then you may be peaking a lower side peaks and not the central large peak. With a lissajous you will see a line at 45 deg (if the gains are the same). It should remain a line as the frequency is swept up to the high frequency for the fine set up at 10kHz. Due to mechanical difficulties at this frequency, the signals will not be stable so setting a peak will be very difficult. The phase then is the important criteria and is easily adjusted as the line is then an Hour Glass shape tiled on its side. You adjust for the smallest cross over point in the middle.

Make no mistake! The Playback HF Azimuth is THE MOSTIMPORTANT ONE IN THE WHOLE SET UP. If this is not set up properly, machine to machine transferability will be severely compromised and ANY tapes bought off the shelf will have severe HF impairment. The only tapes which sound well on it will be recordings done by it. The record head is matched to the playback head setting during the procedure.

Also bear in mind that at the HF end there is no electrical adjustment that will compensate for the mechanical mis alignment. The mechanical side gives the recorder the ability to deal with the high frequencies (with an alignment tape used, deal with the highs recorded on another machine also), the electrical adjustment is to match the highs to the lows. The lows don't move. For the electrical adjustments to work, the high frequencies need to be present in the first place.

As has been mentioned above, don't look for perfection. You will be tweaking all day!!! If the machine meets the spec then the job is done!!!

Now you have a properly aligned, freshly capped machine....... How does it sound????

GPS16
 
Great post! This will be very helpful when I start lining mine up after the Head Block returns from re-lapping. The bass boost is interesting. I don't know if you have that data in terms of dBV or not? Which tape were you using to do the EQ and Balance adjustments? Again, thanks for the through post.

Hi again. I will get around to setting up my other two 707s and see if the bass boost is there too - my ears tell me it is a built-in designed feature of the 707s. The bass boost is still keeping within spec quoted in the manual.

Okay, converted the y-axis to dbV using the following expression (20*log_10 (V/(V0*sqrt(2))). The sqrt(2) is because my measurements of voltage were taken from the scope in amplitude, rather than RMS.
TN3zJx9.jpg

OxVBIjT.jpg


The flat part of the frequency responses from about 200 Hz to 20 kHz is at about 150mV amplitude, or about -19.5 dbV. At 30 Hz the response is up by about 2 - 2.5 dBV about that. For comparison, the frequency response in the manual is for 19.5cm/s (7.5 ips) is "20 Hz to 28 kHz (30 Hz to 24kHz +/- 3 dB)".

To answer GPS, thanks again for the useful comments - now i'm extra glad I got confident about what I was doing with azimuth adjustment! How does it sound? Well, the effort was worth it. I was giggly when doing my first tests - my wife took a photo of me because I was smirking stupidly. It's really hard to describe without getting into a lot of subjective comments since I haven't done a blind A/B yet (I will!), but my impression is.... pretty much everything is better after the work, not just the balance and image position...everything! I pulled out my Michael Jackson Thriller prerecorded album, which has always sounded 'okay' - I don't play it much because it's a kind of pricey tape - it sounded like a different (and stunning) album. The high hat and other effects were zooming around with precision that I hadn't heard before. The tone sounded just right (where as before it was a bit harsh on the top-end). I could go on but you get the idea.

How much of that is from just being mechanically serviced and aligned vs recap I don't know for now (but I will align another 707 and find out). How this sound comes from a 'consumer' deck beats me! Actually, it's probably the most enjoyment I've had listening to any media format!

Okay. So playback is setup. Onto the recording setup.
 
Last edited:
Hi again. I will get around to setting up my other two 707s and see if the bass boost is there too - my ears tell me it is a built-in designed feature of the 707s. The bass boost is still keeping within spec quoted in the manual.

Okay, converted the y-axis to dbV using the following expression (20*log_10 (V/(V0*sqrt(2))). The sqrt(2) is because my measurements of voltage were taken from the scope in amplitude, rather than RMS.
TN3zJx9.jpg

OxVBIjT.jpg


The flat part of the frequency responses from about 200 Hz to 20 kHz is at about 150mV amplitude, or about -19.5 dbV. At 30 Hz the response is up by about 2 - 2.5 dBV about that. For comparison, the frequency response in the manual is for 19.5cm/s (7.5 ips) is "20 Hz to 28 kHz (30 Hz to 24kHz +/- 3 dB)".

To answer GPS, thanks again for the useful comments - now i'm extra glad I got confident about what I was doing with azimuth adjustment! How does it sound? Well, the effort was worth it. I was giggly when doing my first tests - my wife took a photo of me because I was smirking stupidly. It's really hard to describe without getting into a lot of subjective comments since I haven't done a blind A/B yet (I will!), but my impression is.... pretty much everything is better after the work, not just the balance and image position...everything! I pulled out my Michael Jackson Thriller prerecorded album, which has always sounded 'okay' - I don't play it much because it's a kind of pricey tape - it sounded like a different (and stunning) album. The high hat and other effects were zooming around with precision that I hadn't heard before. The tone sounded just right (where as before it was a bit harsh on the top-end). I could go on but you get the idea.

How much of that is from just being mechanically serviced and aligned vs recap I don't know for now (but I will align another 707 and find out). How this sound comes from a 'consumer' deck beats me! Actually, it's probably the most enjoyment I've had listening to any media format!

Okay. So playback is setup. Onto the recording setup.

Very impressive results! looks like you have your machine performing well within original spec (possibly exceeding it), and with a bunch of new and high quality parts within you should be good for another 20-30 years! (with a little maintenance along the way) :-) BTW - what tape did you use to set the bias?
 
Very impressive results! looks like you have your machine performing well within original spec (possibly exceeding it), and with a bunch of new and high quality parts within you should be good for another 20-30 years! (with a little maintenance along the way) :) BTW - what tape did you use to set the bias?

I suspect the specs in the manual and rather conservative, although I don't have test tones on my calibration tape out past 20 kHz so I can't be sure for now; looking at 100 Hz to 20 kHz is basically +/- 0.25 dBV! I will be interested to see when I do some recordings to PC and look at some spectra and see how far the high-frequency extends.

Will report on the recording setup shortly, have gotten half way through and hit a hurdle. I am doing my first attempt with RMGI LPR35 tape.
 
Last edited:
Hi,
Glad you are pleased with the end result of your labours. The 707 is a cracker of a machine. I will be setting up all recorders in my possession to LPR35 for the foreseeable future as this is still available new. Most of my older Maxell tapes are now filled with recordings which will probably be on the shelf and not recorded over. My current stocks are LPR35 3600' and SM911 2500'.

Hope that the Record set up goes OK and the hurdle has been surmounted.

Again, well done!!!

GPS16
 
Moving onto recording setup. Here is an image I made for quick reference that I think is clearer than the similar figure in the manual.
MxHfPDh.jpg



11.4 Dummy Coil Adjustment.


I hooked up the scope via a couple of 15 pF caps (the manual calls for 2-5 pF, but I didn't have anything that small laying around).
t4203ho.jpg

With the deck set to stereo record state the L & R signal are roughly the same (unless your L & R record coils are somehow too different?). Frequency of both channels for me was practically the same at 116 kHz.
ZELOmaF.jpg


You then turn one of the channels out of record state, and then the load presented to the bias oscillator changes and the frequency changes, as the recording head coil of one channel was taken out of the circuit. In L mono state the frequency dropped from 116 kHz to 112 kHz and R mono went to 118. A quick turn of the inductor and both were at 116kHz, same as in stereo recording mode. As I understand, when you switch a channel off of record state a variable inductor is switched into that channel in place of the recording head; we tune the coil to present a similar load to the bias oscillator (see Fig7 in the manual and some explanation section 6.6). So it's just a mono-record precaution. GPS please correct me if I've said something wrong and thanks for your explanations earlier.

Now that seemed simple, and in fact it was done in about 10 seconds per channel..... when I did it properly. A word of caution, those little variable inductors are fragile. I had a feeling something might go wrong since I have never worked with them before so I did measure inductance of the two dummy coils before playing with them. Both were at about 2.5 mH (21 Ohm), which about 2-3x bigger inductance than anyone makes them these days in a 8-10 mm platform, so they aren't easy to replace except with original parts, unless one of us puts in a big custom order. You can see the inductors (in the first image of the post) have square holes. I don't have a square tool so I though I could gently stick a flat screw driver into across diagonals and turn it. Bad idea! That rectangular hole is not in a piece of hard metal or even relatively tough plastic, its a soft ferrite slug. The slug is a cylinder that has that square hole in the middle of it and a screw thread on the outside. This moves inside of a white plastic cylinder and the moving of the magnet in or out of the coil disturbs the magnetic field and adjusts the inductance. Anyway, i stuck my little screwdriver in and gently turned. nothing happened. First problem: just sticking the metal tool into the magnet, shifted the something or messed with the magnet such that the inductor was now 2.0 mH, instead of 2.5mH even with the slug barely moving (if at all!). So don't use a metal tool. The other reason to not use a metal tool? Well the slug is a soft ferrite and your metal screwdriver is much harder. And when I say soft, I mean soft. The slug crumbled pretty quickly under a bit more pressure. I'm sure some of the folk here who work on tuners are having a giggle at me now.

Here you can see the slug is damaged on both dummy coils.
zyfcxFZ.jpg


At that point I decided to dissect the inductor to have a look inside so I know how to handle them better in the future.
Ye7NUmY.jpg

idQLxL8.jpg

You can see the broken ferrite slug (bits with thread) The inner part of the white cyclinder (I cut it in half with my dremel tool) magnet moves in doesn't really have a proper thread but just 3 lines of dots, that the thread 'mates' with. These dots or teeth are also quite soft and I could see that the setups could easily lock up. GPS mentioned to me that sometimes the variable inductors are locked into place with beeswax once setup and gentle formed of heating can be used to get things moving. Anyway, I ordered a spare powerboard to get new coils and also a little set of tools. The square hole is about 1.4mm x 1.4mm. The 1.5mmx1.5mm tool that I got works just fine on the inductors I got on my spare board (and infact the other inductors on my original board, except the dummy coils); it has a long enough 'bit' to contact most of the cylinder before I start turning.

I still couldn't turn the other original dummy coil... tried heat, etc, but would not budge. I suspect they had been glued. At this stage i didn't want to experiment with solvents since the slugs were already destroyed; I replaced both dummy coils with the spares. All of my other coils turn just fine. Piece of cake... with the right tool. One other point is that although the dummy coils have 5 pins, three are not used (only the outer pins on the side with 3 pins are connect to anything - you vary the entire inductor with the magnet adjustment. this is not the type with a sliding contact somewhere etc).

C2VkS0j.jpg


11.5 Bias Trap.
For both L and R channels the voltages at TP301 (Left) and TP302 (Right) were 0.62 V amplitude which is over the 1Vp-p (0.5V amplitude) minimum mentioned in the manual. Adjusting the bias trap coils got this down to 0.32 and 0.36 V amplitude.

11.6 Record Bias.
I pulled out some LPR33 tape and followed the instructions in the manual. Basically you hook up 1 kHz 316mV to the line input and adjust the line input pot (not your source) until you have 450mV at the output when listening to the SOURCE. They are talking about measuring with a multimeter in the manual for this part so i assume they are talking 450V RMS amplitude. Fig. 36 in the manual is really helpful. Basically you start recording and adjust the recording bias VRs (see image at top of post), clockwise from min to max and you will see the recorded playback level got up, and then back down. You are meant to set the bias so that the recorded output is 0.2dB down from he maximum, on the overbiased (clockwise on the VRs from the peak) side. it helps to know before you start turning VRs that 0.2 dB is about 2.3% (if you get a different number you probably calculated for intensity (proportional to square of the signal) instead of amplitude).

To give you an idea of my experience, the output from the recording and maximum upon adjusting the VRs was:

---------- Left ---------- Right
Initial --- 174 mV --- 205 mv
Max ----- 281 mV --- 286 mV

So the VRs were turned until I had a overbias (again, clockwise turn of the VR from maximum recorded output) and 0.2dB drop. I.e.
------------- Left ---------- Right
adjusted -- 273 mV ---- 279 mV

Now the trick is that when you adjust one channel it affects the other a bit. Not too much though. The position of the maxima seems to move so I checked that on each itteration. For example on the right channel my maxima over four itterations of adjustment were 286, 282, 283, 282 mV; corresponding to -0.2dB points of 279, 274.5, 276.5 274.5 mV. Now, looking back at the original value of 205 mV, and the max value of ~280mV, probably anywhere around 275mV +/- a few mV would be fine, right?

11.6 Record/Play Frequency Response.
Setup for this is similar too the last step, except you then drop the input 20dB. I worked in terms of signal/amplitude again so this means a 10x reduction in input signal. I also did a quick check to make sure my el-cheapo signal gen gives same amplitude output when switching between 1kHz and 15kHz. You basically record at 1 kHz and look at the output or multimeter and then jump over to 15kHz input.

At 1 kHz I have 60 mV from the recorded playback on both channels, but 110 mV on the right and 100 mV on the left. You are meant to adjust the record bias (via the same VRs used in 11.6) and so that you have -3 dB +1.5/-1.0 dB at 15 kHz compared to 1 kHz. I wasn't expecting such a deviation from spec. I really expected this stage to be 'tweaking'. In fact, to match up 1 kHz and 15 kHz I can underbias (counter clockwise turn of VRs) so that output is ~256 mV in the setup of 11.6. However, this is a "no no" to underbias below the maximum sensitivity bias value (in my case about ~280 mV output in setup of 11.6). This is my hurdle at this stage.
 
Hi,
the required bias for recording is comparatively small compared to the power required to erase. The dummy coil referred to is switched in to replace the erase head to keep the loading of the Bias Osc the same no matter what the mode. If the load changes then so does the frequency. The frequency need to be stable for the Bias Traps to work effectively. These stop the Bias feeding backward into the audio circuits which would upset the equalisation circuits a tad!!!

The level during the frequency set up is dropped to -20dB so that the tape oxide does not get saturated during the test and set up. The higher the frequency the more energy associated with it which to a large extent renders the bias superfluous. On later machines (and thanks Dolby Labs) the HF content of the source signal is measured and the higher the HF amplitude, the lower the Bias. This is what HxPro is.

Another way of looking at it is the Bass is Infra Red, the Treble is the Ultra Violet. The ultra violet has the ability to do the damage!!!

Also, the HF response falls on increasing Bias but you have to have hit the peak first. Hence the drop by 0.2. If you do the same set up on a Revox you will be setting the initial Bias at the peak followed by a drop of 2dB or 3dB depending on the tape used. Another one of life's strange paradoxes!!!

Keep going. The finishing line is in sight!!!

GPS16
 
Thanks for the clarifications above GPS. The explanations of the circuits in the pioneer manual are much better than I'm used to over in the sansui world, but still there are some blanks. But I'm learning a lot during this rebuild. I will go back once everything is done and look at the manual and circuits again with new eyes and see what else I can learn at that stage!

Yes, I was finding the peak in bias sensitivity in the playback from recording and going past it into overbias, but found that turning further clockwise (as per manual) resulted in increased high frequency at 15kHz (and decreased LF response)! I had to underbias, on other side of the peak, to get LF and HF matched up (or so I thought - read below). And, of course that doesn't make sense at all. And I found the reason everything was back to front.

In 11.6 Record Bias, the manual says to turn the bias VRs "in a CLOCKWISE direction from min to max and then adjust to -0.2dB PAST the max value". I took this instruction to mean that once I hit the peak, I should keep turning clockwise (past the peak) to go into overbias region, and then set to -0.2dB. However, I think the manual is incorrect, or at least unclear.

Turning the record bias VRs clockwise actually decreases bias
. I hooked up my scope again to the record heads to look at the 113kHz bias signal as simultaneously turned the bias VRs just to be sure! Maybe the "turn clockwise from min to max" is correct if they mean VR resistance value, not bias voltage. However, the instruction to turn clockwise and "past" the max value seems incorrect, as this takes you into the underbias region. So from that point of view the manual is saying something wrong. Perhaps it is only unclear if they mean turn clockwise as viewed from underneath the machine (the VRs are adjustable on both sides) but then the figures including Fig.35 all show top (not underside) views.

So in summary I redid 11.6, found the peak of the bias sensitivity (peak recorded playback voltage as your adjust bias VRs) and then turned the bias up a bit more by turning the bias VRs ANTI-CLOCKWISE to get into the overbias region and set to -0.2dB on that side of the peak. For reference I had about 283mV and 294mV for each channel for the input voltage I used. So I overbiased so I had about 275 mV and 285mV.

Okay, now back to 11.7 Frequency response. It turns out that when I thought I was doing it wrong before (going into underbias to match up HF and LF), I actually did it right and was in overbias regime! Since I did 11.6 again I now did 11.7 again.

For my input of about 31 mV amplitude I had about 19mV and 18.4mV on L & R at 1kHz. At 15kHz I initially had 21mV and 23mV. So again, HF was too high compared to LF. But now I know that further overbias is achieved turning the bias VRs anti-clockwise... so I did that until I have about 18 and 17.5mV at 15kHz, which is about -1dB at 15kHz (compared to the new values at 1kHz of 19 and 17.7mV).

I could easily make it about 0dB at 15kHz and in fact might be happier with that because I am now overbiased to the -0.9dB point (as per setup of 11.6, indicating -0.2dB target); the manual says that 15kHz should be -3dB +1.5dB/-1.0dB. But is there really a problem in terms of EQ if I set to about -1dB (which is where I am at now), or even 0dB, for 15kHz? This would get the bias a bit closer to peak sensitivity. Although maybe its not a big deal, or even advisable, to be too close to the peak, with tape variation since you could get into underbias etc etc.?

Okay, basically we are onto 11.8. Recording Level. I guess it should be mentioned in 11.8.1 you need to be in SOURCE mode when adjusting LINE input pot for -7dbV (450mV). They mention SOURCE mode in every other section of record setup but not this one, and it's the only way to setup the input. The 450mV is RMS amplitude, which is what is measured by most multimeters, so just hook up the multimeter to the outputs and adjust the line pot for 450mV at your multimeter with the 1kHz input (I used ~320mV amplitude (not RMS) according to the scope, and then adjusted the line input pot accordingly to get 450mV at the multimeter).

Now when I start recording, I can see the meters are pegged to the max! I had probably set "0" on the meters to 1dbV at some stage, or some similar quick error. So while I had the 450mV input, I adjusted the meter VRs on the playback board so I have "0" on the meters for 450mV SOURCE input. So "0" is actually -7dbV. Oh, got to say again, love the multiturn pots I installed! Another note though is that I did actually buy replacement VRs for the power board and record setup, but I didn't replacement them because I had no problem adjusting these ones, compared to those on the playback board during the playback EQ/balance setup

Once that was done I then started recording again, and then adjusted the record level pots so my recorded playback showed "0" on the meters. The record level was initially at -5dB!

And now I had a rush of excitement, watching those needles sitting at "0", no fluttering or hovering about, but rock solid....knowing that everything setting/alignment upstream, every component, everything.... is doing what it should be, because we told it to!

I haven't made a recording yet. Will do so shortly and report back. Also, I have a few cosmetic things to do to the machine; I'm going to swap the faceplate with a nicer one from another machine, and repaint the black cover. I might not have replaced the belt. Should i just do it? I have the spare belts, but not sure of the point until it breaks since it's so easy to do. But at this stage, although I'm not completely done with the machine, the major learning part of this rebuild for me (a newbie to tapes!!!) has been hurdled. And boy, it feels great just looking at the beast. Got to go.... listen to some tapes now. The thread will continue though, with further updates, anything I learn comparing my rebuilt machine to my other 707s etc etc.
 
Last edited:
Smurfer77 ... you observations regarding 11.6 are VERY interesting. This is very good to know as I would likely have done it completely wrong had I followed the manual. So it sounds like you go clockwise to hit the peak and then counter-clockwise to do the overbias adjustment - correct?

FYI - I tolked to Tom at AMP Services this morning. He said that my Heads were definitely re-lappable and that they will come out just fine. He plans to work on my head block tomorrow. It should be interesting when I get it back to see if the "ills" go away.
 
Smurfer77 ... you observations regarding 11.6 are VERY interesting. This is very good to know as I would likely have done it completely wrong had I followed the manual. So it sounds like you go clockwise to hit the peak and then counter-clockwise to do the overbias adjustment - correct?

FYI - I tolked to Tom at AMP Services this morning. He said that my Heads were definitely re-lappable and that they will come out just fine. He plans to work on my head block tomorrow. It should be interesting when I get it back to see if the "ills" go away.

Great news on your heads!

I assume it is the same on all 707s but will check on my other machines which I will calibrate soon (you can check by either checking the voltage at the record head as you turn the bias VR (if you voltmeter or scope are good for 110kHz) or watching the HF recorded playback... you know you are increasing into overbias direction by the HF decreasing.

Assuming it is the same on all 707s, I would a actually recommend the opposite of the manual; starting with the VR turned all the way to the right (wound out clockwise) before recording. And during recording/playback then turning counter-clockwise to the peak and then turning further counter clockwise into overbias. This way you will avoid any unnecessary large voltages in the circuit etc by avoiding the starting scenario of bias voltage set to maximum (I don't think there is any real danger in this case, but that is standard procedure with adjusting VRs).
 
Last edited:
Hi,

to adjust L301/302 effectively, you would need to be able to measure distortion. If you plot the Bias against response at 10 kHz you have already seen the peak followed by the drop on the over bias side. If you superimposed the distortion on the same plot you would find that the distortion is around the minimum at the over bias side. If you do not have this specialist measurement gear then the L301/2 adjustment is best seen as an initial adjustment and a factory only one.

On some equipment I have serviced, there are some resistors which are marked SOT. These are usually components which are "Selected On Test" on the PCBs first power up (may be on a specific test rig to do so). Bear in mind that this coil will affect 4 modes of operation at the same time (2 speeds and 2 equalisations). You will be adjusting a compromise on all 4.

If you don't have the correct test gear then leave them alone. You may do more harm than good!!!

GPS16
 
Hi,

to adjust L301/302 effectively, you would need to be able to measure distortion. If you plot the Bias against response at 10 kHz you have already seen the peak followed by the drop on the over bias side. If you superimposed the distortion on the same plot you would find that the distortion is around the minimum at the over bias side. If you do not have this specialist measurement gear then the L301/2 adjustment is best seen as an initial adjustment and a factory only one.

On some equipment I have serviced, there are some resistors which are marked SOT. These are usually components which are "Selected On Test" on the PCBs first power up (may be on a specific test rig to do so). Bear in mind that this coil will affect 4 modes of operation at the same time (2 speeds and 2 equalisations). You will be adjusting a compromise on all 4.

If you don't have the correct test gear then leave them alone. You may do more harm than good!!!

GPS16
I think I'll leave mine alone! I do have a distortion analyzer function on my audio toolbox, but I don't know if it will do what is needed and I don't feel like I'm up to that level of tweaking - unless I can't get good performance any other way. I'm hoping that once my heads are re-lapped and aligned most everything else will fall into line pretty well - fingers crossed!
 
I don't have a distortion meter yet (although I will soon, am building up the kit for stereo FM alignment). However, my scope does have a real-time FFT. I can probably see the minimum in the distortion via the amplitudes of the FFT harmonic peaks for the time being. (I have also been playing with hooking up my scope to PC and using matlab code to calculate the distortion from the data, but that is in 'alpha' stage now). Anyway, I will just see if I can see measure the distortion via the real-time FFT as I align my second 707, probably this weekend. (Will just play with measuring it in the bias setup, rather than start playing with the peaking coils.... don't want to brake anything that aint broke.)

Oh, last night I was thinking that at the beginging of play setup to 450mV amplitude. However, they mention confirming the 450mV by multimeter, not scope etc, and most multimeters present V_rms. So they mean RMS, right? If someone knows better please let me know. So I reset the gain/balance. Then I've redone the playback setup because the gain adjustment affects EQ, and the relevant last parts of record setup affected too.

No surprise that I have gotten much quicker at the process. And starting to know my beast. I wont even make a graph this time, it's just embarrassing how good it is. I do a quick check at 1kHz, 10kHz and 20kHz before adjusting then good to keep in mind that whatever you do at about 10kHz will be exagerated at 20kHz. For example, if 20Khz is up (down) by say 20mV, you adjust 10kHz down (up) by a a fraction of that, maybe ~3mV.

Two other things, beware there is the mains socket wired just were your hand might touch, as you adjust the dummy coils in particular. I am pretty careful but I put some tape over the exposed joints while I was working.

Another thing happened that confused me for a moment. When I recalibrated this time I used multimeters instead of scope to measure the output. One channel seemed to have low high frequency output but it just turned out that the el-cheapo multimeter I was testing didn't work properly for AV voltage for frequency >3kHz (claimed to 100kHz). So it looks like I am going to buy a second Fluke after all. It was actually easy using the Fluke than scope partly since the multimeter has a slower response time than the scope so it's easier to read an average voltage at high frequency when the levels jump around a bit.

okay, off to listen to some more tapes on the 'new' machine!
 
When measuring audio voltages, it's always preferable to use an analog AC millivolt meter.

http://www.museu-tecnologia.com.br/wordpress/?p=239

To be clear, I'm not just doing it visually on the scope, but using a real-time measurement calculation of the digital scope. On that calculation I can play with the response time, so it's not too bad, but my dvm does a better job. But boy oh boy, I can see why the 2 channel analogue readout would be nice - lining those two needles up when balancing L & R channels must be a treat... and i bet those do a nice job of averaging the flucations, which is the issue with my scope. I love my vintage military Fluke DVM and was thinking of getting a second, but am definitely getting something like this. thanks!
 
Last edited:
Hi,
we use ultrasonic cleaners at work. If you put your finger ends in the solution it feels like pins and needles in your fingers. You can see the dirt from under your finger nails come off like smoke. As a test we used to put some kitchen foil into it and the scrubbing action should make holes. Wonderful things to play with!!

GPS16

Finally got around to sticking my fingers in the ultrasonic cleaner. yup, like smoke coming out from under my fingers and feels like pins and needles. my wife tried it, and it didn't work.... i guess she keeps her nails cleaner than me.

I've starting ripping some prerecorded tapes to PC. The spectra shows response goes well past 20kHz when I sample at sufficiently high rates - my ADC does up to 192kHz - so I guess my heads are physically still good and setup well enough. Nice thing is my ADC (Focusrite Forte) also works as a DAC and an analogue output so I can use the same device for playback or recording back to tapes.

As for further work on the 707s, aside from calibrating and fixing up the other two 707s I think I will repaint the black lids of all three... they are all no perfect. Looks like the paint has small specks in it, so I will try to replicate that original finish.

Is it worth making a back-up copy of my calibration tape once I have two machines calibrated? I don't mind buying another calibration tape in the future should I need one but that assumes the tapes are still being made.
 
Howdy,
there is no point in making a backup copy of your calibration tape. As good as a 707 is, it is no where near quality enough to make reference recordings!!! Store the reference tape on it's edge, away from direct sunlight, away from magnetic fields and in a cool dry place.
If you are happy with responses past 20kHz then the reference tape has done it's stuff!!

Whilst messing with the US Cleaner, was going to put something else into it but bottled out!!!!

GPS16
 
Howdy,
there is no point in making a backup copy of your calibration tape. As good as a 707 is, it is no where near quality enough to make reference recordings!!! Store the reference tape on it's edge, away from direct sunlight, away from magnetic fields and in a cool dry place.
If you are happy with responses past 20kHz then the reference tape has done it's stuff!!

Whilst messing with the US Cleaner, was going to put something else into it but bottled out!!!!

GPS16

Fair enough. Still trying to get hold of a 7" reel with the big hub to reduce stress on the tape. But in the meanwhile I've just added a very generous length of leader tape at one end, so the tape gets wound at a bigger radius, and a bit of leader tape to the other end so that the magnetic tape is well past the heads when I reverse direction.

About the ultrasonic cleaner... that would take some serious guts.
 
Back
Top Bottom