Scott 222C

[NOSValves]

I'm not sure if that dial on the 667 translates into actual plate voltage but it might. But regardless the test results will be completely flawed. The tube will have completely different characteristics at its true operating voltages. Any of the vintage tube tester that you can most likely purchase affordably as a DIY solder slinger is not really accurate for anything but very basic safe to use or not safe to use results. They in no way will match tubes for Gm or current. To get accurate results the tube has to be tested at true operating voltages and current.

Heck I have 4 Hickok tube testers, 1 B&K and a EICO 667 I still know I can't truly match tubes :tears: But I do not sell tubes so I'm not really concerned with it. I match them in the amps I service and this method is more accurate then even what the tube sellers use since the tube is at absolute operating conditions.

Any self respecting tube seller at the minimum will have what is called a Maxi-Matcher which has the ability to apply 400V to the plates, screens and a variable negative bias voltage setting.

But any way what you need to do is first find a set of your available tubes that reasonably match in your amplifier. The 4 you have are grossly mismatched. Especially the one you have in slot 3 it was only drawing 7 mA of current while the rest were at 13 to 15. Ideally and I'm not sure how many of these tubes you have but I'd try keeping the two that drew 13 mA and try to find two more that are closer. The one that drew 15mA would not be to bad to use with the two 13mA tubes. But if you have some closer why not use them.

Oh before I forget and I don't want to depress you but the 667 is just a basic emission tester it doesn't even give you mutual conductance scores. I own one because then can test good and bad on so many different tubes and they have a really good leak test function. But honestly the tester is completely useless for matching tubes. I know it says "Dynamic Contuctance" but that's just a fancy way of saying "Emision Tester"

I think I mentioned a while back on the forum about buying these russian tubes on eBay for a song...Now you know why I don't advise for someone that isn't in the market for say 100 of them so they can weed through and match them. Just pay the $50 from a reputable tube dealer and get a matched quad. Way less headaches. Sometimes getting something cheap isn't the best way to do things.

 

[thefragger]

Alright, I'll try finding a close quad from the bunch that I've got. So I basically put the set in and balance the voltages between the sockets, then measure the current draw from each tube, and I want a set that will be as close to each other as possible in terms of current?

And then assuming that I've found a set of four, what's next?

 

[NOSValves]

Yup balance the Negative voltage to all 4 tubes. I'd install the 2- 13 Ma you already know match each other then start plugging in the other you have finding the closest 4 you have. If you want to make this quicker you can install those 50 Ohm 50 watt resistor from the 0 speaker tap and a 16 Ohm speaker tap to place a load on the amp. Then you can hot swap the tubes to make a quicker proccess without constantly turning your amp on and off. Use glove on the hot tubes!

Once you have a reasonably close match then we can figure out how you want to accomplish the biasing. Many ways you can go with that from simple to complicated.

 

[thefragger]

Alright, I found two pairs;

pair 1: 13.23mA / 13.10mA
pair 2: 15.80mA / 15.84mA

Do I have to use a quad or will two pairs work?

 

[NOSValves]

Alright, I found two pairs;

pair 1: 13.23mA / 13.10mA
pair 2: 15.80mA / 15.84mA

Do I have to use a quad or will two pairs work?

Yes that will work.

Now these scores were with all 4 tubes receiving close to identical negative control grid voltage correct?

Actually if you just want to do the minimum adjustment to get the bias up then it will be best to install a 13ma with a 15ma in each channel then balance them even. Then if we just do something to lower the grid voltage both channel will end up with similar current.

If you're going to add a fully adjustable bias scheme then you can run the 15's in one channel and the 13s in the other since you will be able to raise them individually.

How do you want to tackle getting the bias current up? I've went this far so I might just as well help figure out the rest.

I'm signing off for the night just post how far you want to go and I'll cook something up in the morning for you.

 

[thefragger]

Wonderful news! When my financial aid comes in from my school I'll nab a proper quad of Sovtek EL84Ms from Jim McShane--I payed all of first semester out of my shallow pocket and with much family help!

The bias for all the tubes (pin 2 to ground) is -18.88V, +/- 0.01V.

I'd love to hear music as soon as I can to enjoy all the work I've put into it, but I like the idea of being able to adjust the bias of each tube independently and since it's already on the bench I might as well try for the independent-biasing functionality. I like how the 299B has one pot (BIAS) as a voltage divider and the next one (BALANCE) to even out the two tubes--simple and straight forward. Would it be as simple and putting in four multi-turn potentiometers, one for each tube, to directly and accurately dial in the bias voltage?



Craig, you're a wealth of knowledge and I cannot thank you enough for helping me out so so much with this rebuild--I owe you more than a couple beers, that's for sure!

 

[NOSValves]

All my expeeince is from hands on. It started with helping my father at a ripe old age of about 6....My dad was a appliance electronic repiar man by day and a Ham radio restorer by night. I grew up in very rural northern michigan where back then the snow was very high all winter and we had one single lousy TV station... So I spent my time helping my Dad at night. I never did take to a ham gear but ended up loving Stereo equipment. So here I am I turned my hobby into a business 10 years ago and ruined my hobby in the process. But I do enjoy what I do..

Don't waste your financial aid on buying another quad of tubes! The ones you have will work just fine.

You can use the amp right now. But the voltages will be high but with the reduced current the tube will be safe. The amp is just not going to sound its best. If you want we can just insert a single resistor and another 50uf at 100V cap for the time being to get the bias current up and then revisit this when you have the funds to source the parts.

If you want to be able to completely bias the amp then I suggest you simple mimic the circuit already designed for you on the 299B. The 222C is really just a stripped down version of the 299B anyway! Once you have that circuit in and working or close to working if you want some help tweaking its range of adjustment just let me know. Also you may want to install 4 bias test points and 10 ohm biasing sensing resistors one for each output tube so you can bias the amp from above the chassis without removing it from your system. Keeping them tubes properly biased and balance will extend there life and lower the over output section distortion. Another big benifit is you won't need closely matched output tubes to achieve minimum distortion (but you still do not want grossly mismatched).

 

[thefragger]

Wow, that's a very cool story--I got into electronics from starting at a hobby level some 16 years ago when I played with discarded VCRs and was winning science fairs with crystal radios. From there I got into robotics and eventually repairing my dad's old audio gear for fun and taking residential electricity in high school. Hell, my hobby got me a job at one of the world's foremost hospitals working in the R&D wing developing cutting-edge prosthetic devices, all with no profession training! While now I'm perusing a diploma as a microelectronics technician while I'm also working on my degree in mechanical engineering (mechatronics major), I still love learning and getting myself into situations and projects that are way over my head.

I printed out the 299B and the 222C schematics' output sections and you're very right--I didn't see it before but side by side they're damn-near the same! I played with some simulation software and I think I've got some preliminary values for the bias adjustment setup which varies the voltage from -2V and -23V. I swapped out the original 330k resistors for 150k ones and the 15k bias cut-off resistor with a 500ohm one. How does it look?

​

 

[thefragger]

Ok, that 500k resistor dissipates nearly 3.5W (with the full 44V through it)! I changed the bias pot to 100k and the cut-off to 10k which looks a lot better in the simulation while keeping the adjustment range the same as before.

Also, for putting terminals in, I'd need to break the current cathode-to-ground connection in order to put the 10ohm resistor in, or do I add it to the circuit as-is, and then there's one terminal for each set of tubes or one terminal per tube?

 

[NOSValves]

Well that a nice effort but I'm not sure why you feel you have to redesign what was already designed for you? That is a DIY trap... many guys think you have to modify just for the sake of modifying... In many cases the improvement gain in one area of a mod is a derogation in another. These engineers of the 60's actually had college degree's garnished from electronic tube theory! Go the college today at you learn just about nothing about tube equiptment circuits. They forgot more about how to get the best balance out of these products then most of us today know!

For the most part you simply need to mimic the 299B schematic for the control grid voltages. Do not change the value of the 330K's because you will skew up the frequency cut off locations. If you change the 330K value you would also have to change the coupling cap value to correct the resistance change. You already have some of the best PIO caps available of the proper value in place so why force a change that is not required..

All you have to do is install the two pots in front of the balance pots. Nothing more, nothing less the value of the resistor to ground will be adjusted to get the proper range. Also you would never want the low cut off adjustment to reach a voltage as low as -2V if someone accidentally messed with those controls and turned it down that low the tube would run away and it could take out a transformer. You never want that voltage to get below -10 and usually negative -12 to -20V will do the job just fine on a 222C.

 

[NOSValves]

You only need 10K pots....25K max..

 

[thefragger]

It seems I mis-remembered what you said;

Dig up the schematic on the 299B and you have the layout for the controls. You will have to fudge a few resistor values to get the range of adjustment you need. Basically you want -12 to -20. Do all the fudging with the 5AR4 out of the amp but all other tubes present.

I'll get that fixed up!

 

[thefragger]

With the 330k resistors in place, I can only get the bias down to -16V, it seems. With the 10k pot and a 25k resistor I can get the bias varied up to -11.25V, but those 330K resistors are keeping the bias floor at -16V. In the 299B schematic they used 220K resistors in that position, and if I do the same, the variance range changes to -13.7v to -19.75V.

If I use a 10k pot and 15k resistor for the bias control, and swap out the 330k resistors for 220k parts, I can get the bias range to be -11.57V to -19.75V in the simulation; I can't get that -12V to -20V range with the 330k resistor in there.

Also, I'm a bit confused about installing the 10ohm resistors. since the cathodes are grounded already, do I break the ground connection and install a 10ohm resistor between each tube's cathode and ground? That's the only way I can see it to get current flowing through them rather than paralleling them with a short to ground.

 

[NOSValves]

With the 330k resistors in place, I can only get the bias down to -16V, it seems. With the 10k pot and a 25k resistor I can get the bias varied up to -11.25V, but those 330K resistors are keeping the bias floor at -16V. In the 299B schematic they used 220K resistors in that position, and if I do the same, the variance range changes to -13.7v to -19.75V.

If I use a 10k pot and 15k resistor for the bias control, and swap out the 330k resistors for 220k parts, I can get the bias range to be -11.57V to -19.75V in the simulation; I can't get that -12V to -20V range with the 330k resistor in there.

Also, I'm a bit confused about installing the 10ohm resistors. since the cathodes are grounded already, do I break the ground connection and install a 10ohm resistor between each tube's cathode and ground? That's the only way I can see it to get current flowing through them rather than paralleling them with a short to ground.

I have no idea what that program is doing or what values your are using... Those program IMHO can be flawed for some circuits... The first thing you probably do not know is the proper amount of current that will flow through the circuit so that value is probably way off and the current has everything to do with the results. First off the 299B utilizes a 5K bias adjust pots but a 10K or 25K will work just fine also.... It also uses a 15K resistor to ground from one leg of each pot... This 15K value is how you adjust the low side of the adjustment. The higher the value the higher the low point of adjustment will be. The lower the value the lower the low adjustment point will be with minimum effect on the maximum voltage. If you want more or higher negative voltage you simply use a spot closer the the negative voltage source in the RC network. This is a really simple circuit and it seems like your trying to make it complicated. I'd suggest you quit over analyzing and just start doing you'll learn more that way rather then rely on that program to do your thinking for you. As long as you don't fire up the amp with the 5AR4 in you can't hurt a thing with this circuit.

As far as the 10 Ohm cathode resistors yea its a pain on the 222C because they used those cathode tube socket eyelets for a few components ground tie points. You have to make a new ground point to move every single connection now on the cathodes and then insert a 10 Ohm resistor between the cathode and the new ground points for bias measuring... 10 Ohms makes it simple to set the bias. If you want 23 mA you set the cathode to read .23 VDC. To put it simply for every MA the plate draws in current the cathode draws the same MA from ground.


By the way changing the value on that grid resistor will have very little effect on the voltage the tube expereinces the current draw via the control grid with a good tube is really small.

 

[NOSValves]

Just so you maybe feel a bit more comfortable I'll tell you I have probably installed this exact circuit on 200 or 300 222C's in the last ten years trust me it will work just fine. In fact if you spent enough time searching old posts on the HH Scott forum you would find where I posted a rough schematic of this circuit about 9 or 10 years ago. I install bias controls and test points on just about every integrated amp of any brand that it will work in.

 

[thefragger]

Thanks for that bit of reassurance--I never doubted you, but I just wanted to see what the parts did and how their values influences the circuit--you said above that I might have to adjust some resistors, so I was looking for which parts to play with to get my range pegged.

Dig up the schematic on the 299B and you have the layout for the controls. You will have to fudge a few resistor values to get the range of adjustment you need. Basically you want -12 to -20. Do all the fudging with the 5AR4 out of the amp but all other tubes present.

I basically hit the basement at 3p and wrapped it up around 6.30p yesterday, with a couple interruptions, but it's all installed now. I added two bits of tag board to hold the 10 ohm resistors and the new star ground point as well as installing two 10k potentiometers that I had on hand. I didn't have any 15k resistors but I was able to parallel three 47k resistors to get 15.67k as a stand in for now. I don't have any tip jacks on hand so those will be installed later for measuring the bias from the top plate.

So now I have two variables per channel that I can play with--the bias voltage supplied to each pair, and how the voltage is distributed between the two of them. I guess now all I have to do is get the bias set up correctly with this new arrangement and I can start rocking out?

 

[NOSValves]

Yup sounds like you getting close to rocking out.

I didn't think you had parts to jump right at this so you may have went a bit further then I anticipated. I should of given you some of the known differences before you actually soldered things in place.

Here is what you should do...

Step one

Change the voltage source point feeding the two bias adjust pots. Take it right back to the junction of your 33 ohm and first 20 ohm if the bias/heater supply R/C network. Place no dropping resistor or filter cap between the source and the pots.

Step two...

Source two electrolytic caps something like 10.20,30,40 or Ideally 50uF at 50V or higher.

Install these on the wiper of the two new pots (wiper is the center lug) remember this is negative voltage so the caps go reverse polarity. Negative lead to the wiper positive to ground.


Now fire up the amp with the 5AR4 out and get us a base line. Blance the voltages again on all control grid (pin 2) then adjust the both new pots all the way for maximum voltage and record it for both channels. Then go all the way the other direction and record the low voltage reading.

If everything is right we should have plenty at the high side maybe even more then we need which is fine. The low side probably will not be ideal and you will have to change the value of the 15K resistor.

I've never seen two of these amp that ended up identical in a row.... the value of that resitor to ground is always slightly different. Then factor in all the different models I do which are all stightly different I've never really recorded what values i tend to use. I always tag a variable resistor in place for the 15K to ground adjust it for proper range then read its value and put a perminent one in when the actual value is known.

 

[thefragger]

Ah, okay. Doesn't sound too hard to make those changes, I've got a pair of 47u/63V Panasonic FC caps right here (and a pair of 47u/50V Nichicon PW caps). Swapping out that 15k part might be the trickiest, but doable--I've got some small 10-turn pots I can put in there to play with, too.

I'll get the changes made, the parts installed, and I'll get those values in a bit...

I've got craploads of parts (I really need to clean my bench!)

​

 

[thefragger]

Alright, here are the voltage ranges for the two sets of tubes, with their control grid voltages balanced;

tubes 1-2: -13.26V to -21.20V
tubes 3-4: -12.47V to -21.15V

I'd say that that's pretty close to the target range of -12V to -20V! That's with the 15.67k composite resistors I had in there.

 

[GreatTone]

This has been a really interesting thread. I was just wondering the logic of putting 10 ohm resistors on the cathodes. On my Deluxe Reverb I use 1 ohm resistors, so the conversion from mV to mA is direct -- no moving the decimal. A small thing I guess, but I was just wondering why not go ahead and use a 1 ohm?