After my previous adventures a few posts back, checking the electronics of the playback board, I concluded that my differences between left and right channels on forward playback were due to different efficiency/response of left and right sections on the heads. So, I've caved in and
ordered a calibration tape. I really don't see a way around it beyond making an electromagnet device to simulate the tape (that doesn't sound too difficult, but the time/expense start to make the tape look like a good deal).
The original Pioneer STD-154 calibration tape is not easy to come by, and there would be little assurance that the tape still has the calibrated fluxivity, which it wouldn't, depending on use and storage. So I didn't want an old STD-154. I started looking around at new calibration tapes that would be suitable and also had some messages with GPS16 & decided to go with the
Magnetic Reference Laboratories (MRL) 21T204, which is a 1/4" wide, 7" reel, 7.5in/s (190 mm/s) NAB EQ tape with 250 nWb/m reference fluxivity. I put in an order with Chuck at MRL via email and used their secure webform to give my card details and I forget to include my security number. Chuck gave me a call with an hour and popped the tape in the mail the next morning. They seem like nice guys there. The delivery was extremely quick, but hey, I'm only 20 miles away
. Here is a picture of the tape in it's box, which came packaged within another box with styrofoam peanuts.
The tape is a 2-track/'half-width' type (L and R signals each taking up approximately half of the 1/4" tape width and the sound runs only in one direction, as opposed to the 4-track tapes I'm used too on which you have L_fwd / R_rev / R_fwd / L_rev across the tape width for bi-directional play). The tape comes wound "tails-out". Being new to tapes I only learned what that means a couple of weeks ago; for those who are not sure it means that the tape is wound with the end (tail) of the program hanging at the outer part of the reel. E.g. if you played a 2-track program onto a take-up reel, then the tape is 'tails-out' on the take-up reel. The purpose of this is to turn possible 'print through' of particularly dynamic signal from one layer to others (the program on the tape is a magnet after all, the stronger sections can magnetize other sections) from pre-echos, into post-echos, which are less noticeable for a few reasons. Apparently all 2-track material should be wound tails-out. I'm sure there could be some debate on the need to do this, but let's skip that here. Anyway, all of this is moot for 4- track materials since there is no 'tail' since side one and side two run in opposite directions.
Okay, sorry about that. Moving on.
First thing I did was demagnetize and clean my heads - would be a pity to partially erase my calibrated tape with magnetized playback heads. With the intention of
calibrating the fwd playback head at first I popped the tails-out calibration tape onto the right motor and wound it onto a take-up reel on the left motor so that the tape is 'heads-out' (can we say that, or it's 'tails-in'?) on the take-up reel, ready to play forwards.
Now, the original Pioneer STD-154 tape has content like this:
All of this can be used but the key things are the first tone at 700Hz at 0dB, used for setting the overall gain, and the -10B 700kHz and 10kHz tones which the points the service manual focused on for EQ calibration (see my previous post about this procedure).
The MRL tape I ordered is slightly different, with details in Table 2 in this pic:
BTW the MRL tape comes with a nice calibration sheet:
The main difference between the MRL tape and the pioneer tape is 1000kHz at 0dB instead of the 700Hz. I don't see any problem with using 1000kHz to set the overall gain instead of 700Hz, assuming the response is relatively flat within the region (which it should be). Regarding the EQ calibration, I used either 500Hz or 1kHz -10B (the manual calls for 700Hz) and 10kHz at -10dB. Again, this actual procedure is outlined in
post #23 with the difference being that now I used the tape, instead of a signal generator as my source. If anyone has questions about it though, do ask, because it's not that straightforward based on the manual IMO.
Immediately on the first tone I could see the difference in the L and R forward head pickup (recalling that my electronic circuit was L/R balanced in post 23). L output was ~500mV out and R was ~750mV (and recall that it is my right channel sounding too loud - no surpise). VRs were adjusted to ~450mV on both channels, then I did the EQ adjustment (you will have about output of ~150mV at -10dB, as described in post 23). My EQ was way off on one channel (down to 100mV at 10kHz) and it was easy to fix (although this resulted a couple of calibration iterations (going back and forth between playback gain (11.1 in manual) and EQ (11.2 in manual), since the VRs involved affect each other in-circuit, as mentioned in post #23.)).
Forward playback is calibrated AND sounds great in terms of balance and tone. I do note that there is a tiny midbass (observed at 125 and 250Hz) hump when output at 500Hz/1kHz and 10kHz are used as calibration points. It's a small hump (10mV) and probably a part of the gain circuit. (I did do a minor azimuth adjustment - see later in this post for details).
Okay, so I wound the calibration tape back onto the take-up reel, swapped reels (and flipped the reels over too of course). While fwd calibration was very quick, I struggled a bit with the
reverse playback calibration. Setting the gain such that 1KHz (0dB) gave equal L/R output of about 450mV, made it difficult to have similar L/R output voltage after the EQ calibration. The best I could do (after EQ is calibrated between 1kHz and 10kHz) was 10mV difference between L/R at 1kHz (-10dB) gives - right channel was down. No amount of itteration in gain and EQ adjustment sorted this out. I understand that in theory anything is possible, but in practice I came to a dead-end 'winding out' a VR. Okay, so maybe I had a brain fart, and if i sit down and do this calibration again I will work it out, but I think L and R are really quite different in their efficiency and I wonder if some head adjustment is needed. I will have a another crack at the calibration again, but (a) forward playback is outstanding now, both in measurement and listening tests and I can live with the slightly off reverse playback for now and (b) i will re-calibrate reverse playback, but not until after I replace the VRs (see notes at end of this post).
I noticed that the azimuth adjustment was slightly out for both directions. I could not see it looking at the tape visually, but i noticed a phase difference between L and R output on the oscilloscope, so I gave it a slight adjustment. And, I mean slight. I limited myself to adjustment that didn't move the tape out of the head wear groove. Honestly, I would like to get my hands on some unworn heads (how expensive is relapping btw?) and do all head adjustments properly (I'm not as confident in factory adjustment as many others might be, based on what I'm seeing so far). If you want a surprise, grab the back an envelope, work out the distance along the tape for each cycle of a 1kHz signal assuming 7.5ips, and then work out the distance along the tape corresponding to, for example, a 45 degree phase difference (1/8 of a period/distance of single 1kHz cycle on tape). Then take that distance, assume 1/8" between tracks, and apply your inverse tangent and see how small the angle is that results in such a large phase difference. As a result I'm thinking visual inspection of azimuth is insufficient. But again, I'm new to tapes.... so don't pay too much attention to my ramblings if you know better.
A few notes:
- It is nice to have the 2nd 1kHz 0dB tone at the end of the MRL tape. You can adjust the VU meters using this after calibration, instead of winding back to the same tone at the beginning of the tape.
- The "duration of tones" are noted as 10s each (it might actually be slightly shorter due to gap + spoken voice announcing frequency and level - i didn't measure beyond using tape counter so can't say for sure). While this might be long for an experienced calibrator, I found it maddeningly short.
- This brings me to a final note: the VRs are my machine have to go. Part of the difficulty with the 10 second tones is that the VRs are extremely fiddly. The 'wipers' in my VRs also have a fair amount of static friction (which makes tiny adjustments rather random) despite my best efforts to clean and lubricate. The most maddening aspect is that resistance value shifts as I remove the screwdriver force - and remember, you are trying to do these adjustments rapidly to save rewinding the tape. Replacing these VRs with new 'multi-turn' VRs will be high on my priorities list for electronic refurb.
- And, I do think I don't want a tape with very long tones, as it's more rewinding and fast forwarding time to jump around between tones required for the multiple iterations of calibration.
Finally, I have a question. Is the STD-154 also a 2-track tape? And, does anyone know if the tapes are really half-width tracks, or 1/4 width tracks (with reverse tracks being blank)? I want to know because this affects what you would expect in terms of output during azimuth adjustment (which is rotating the head in the horizontal plane, parallel to the floor/ceiling).
