[JoshK]

Re: tube other than 12au7.....yes indeed, 'cept that the 12bh7 is basically a variant of the 12au7...basically the same tube but better, or at least operates similarly.

On the schematic, the A/B/C/D points connects to the transistor fed grid bias pot.  I've never seen this scheme before, but I haven't read many pentode style amp schematics before.  I guess this is some sort of quasi ccs scheme.   Weird to see a ccs on the grid. 

The input bias pot scheme I have seen before a couple of times....but I am not sure how you set it, as I imagine the gm of the tube comes into play. 

[woodsyi]

gitarretyp,

Don't you want to check out other 12au7s before you do all this?  I got a bunch of OS and NS 12au7s.  I would be glad to loan them to you if you want to try different tubes. PM me if you want.

[Bill Baker]

Don't you want to check out other 12au7s before you do all this?

 Good idea, I thought he already did this but after reading back through, I see it was the power tubes. I don't think this is the reason only because both channels are acting the same but it is definitely a very cheap first step that could potentially save a lot of hassle.
 Good call woodsyi.

[Bill Baker]

Re: tube other than 12au7.....yes indeed, 'cept that the 12bh7 is basically a variant of the 12au7...basically the same tube but better, or at least operates similarly.

 My apologies, I should have stated ".... tube other than 12AU7 or variant...."

[gitarretyp]

Firstly, i am comfortable poking around inside the amp. I've built a few tube guitar amps and modded most of my other electronics.

Second, the A,B,C, and D points are as JoshK surmised. Bias is measured across the 10 ohm cathode resistor, and the suggest bias current is 55mA. I've tried running them between ~38mA and ~60 mA with little perceived difference in headroom. I'll try experimenting more with that tomorrow afternoon. I'll also get the OPT's primary impedence.

Also, i have not tried any other input tubes because the only other preamp tubes i have on hand are 12ax7s and 12at7s. It currently has 4 JJ/Tesla KT88 and 4 Jan Philips 5814A. The previous owner claimed the JJs have ~100 hours on them and the Jans maybe 1000 (he seems to be a very honest seller, so i do trust him).

Woodsyi, let me look into a couple of other things tomorrow, and i may take you up on the offer which is much appreciated.

[Bill Baker]

Firstly, i am comfortable poking around inside the amp. I've built a few tube guitar amps and modded most of my other electronics.

 Good to hear. I was not implying anything. Simply did not want to see anyone get into unfamiliar territory and hurt themselves.

 Keep us posted on your progress. I am very curious as to what's going on with this amp now.

[gitarretyp]

Firstly, i am comfortable poking around inside the amp. I've built a few tube guitar amps and modded most of my other electronics.

 Good to hear. I was not implying anything. Simply did not want to see anyone get into unfamiliar territory and hurt themselves.

 Keep us posted on your progress. I am very curious as to what's going on with this amp now.

No offense taken. I understand that you need to careful recommending working on electronics (always better safe than dead ).

I'll let everyone know what i find out tomorrow.

[gitarretyp]

I was bored this evening, so i measured the impedence of the OPT primary. It's ~145 ohms across the primary and about half that from the primary to center tap.

[gitarretyp]

Alright, i have some numbers:

The voltages on the cathodes of the input tubes balanced at 8.96V and 8.85V for right and left channels respectively. This is over twice what is shown on the schematic?

The only other voltages that were way off were the plates of the second preamp tube. Pins 1 and 6 of the right channel were 255 and 287V and 1 and 6 of the left channel were 220 and 327. According to the schematic, these should be at 280 and 257V for pin 1 and 6. The cathode voltages on these tubes were 174 and 170, right and left.

The B+ coming off of R5 was 317V, about 10% high.

Not listed on the schematics but possibly of importance, the plate voltages on the output tubes were 483 on all four tubes.

If i didn't mention the value, it was within ~5% of the schematic.

After i checked everything and re-balanced the input tubes, i listened again. Now, the right channel is roughly 3db louder than the left .

Any further input would be great.

[gitarretyp]

By the way, for those not familiar with this model, it's the amp only version of the aq-1001, predecessor of the aq-1001 dt

[Bill Baker]

After i checked everything and re-balanced the input tubes, i listened again. Now, the right channel is roughly 3db louder than the left

 After "re-balancing" the input tubes, did you double check the bias of the output tubes? Or is this what you meant.

 This is an odd one. Obviously it is hard to diagnose a tube amp over the internet as the mind is not quite focused on it in a way that everything adds up properly.

[eico1]

I wouldn't use speakers in place with a SPL meter to evaluate amplifier performance and I definitely wouldn't assume I could reproduce a speaker manufacturers specs.

Use a resistive load and a DVM for the amp.

Unless you are an acoustician, leave the SPL meter at radio shack!

steve

[gitarretyp]

After i checked everything and re-balanced the input tubes, i listened again. Now, the right channel is roughly 3db louder than the left

 After "re-balancing" the input tubes, did you double check the bias of the output tubes? Or is this what you meant.

 This is an odd one. Obviously it is hard to diagnose a tube amp over the internet as the mind is not quite focused on it in a way that everything adds up properly.

I balanced the voltages on the input tube cathodes and double checked the output tube bias when finished.

I'm leaning towards the preamp tubes being out of whack since the voltages on the left channel's plates were so far from spec.

I wouldn't use speakers in place with a SPL meter to evaluate amplifier performance and I definitely wouldn't assume I could reproduce a speaker manufacturers specs.

Use a resistive load and a DVM for the amp.

Unless you are an acoustician, leave the SPL meter at radio shack!

steve

The 3db difference was based on inputing a 100Hz tone and measuring output voltages on the two channels. One was almost twice the voltage as the other (ie, 3db).

[gitarretyp]

Well, today i played some more with the input tube cathode bias pot. I fed the amp a test tone and matched the output level of the left channel to the right. With the left level matched to the right, the left input tube's cathodes read ~8.8V and 9.2V and the plates on the second input tube on the left channel read 242 and 296 (closer to the schematic and closer to the safe maximum plate voltage for a 12au7). After i did this, i re-listened and the channels match and the output level is improved. It's no longer clipping at too low a level. Yay!

Now that i have the amp fixed, could someone tell me a way, if possible, to increase the gain from the ~20db it is now to about 30db (or equivalently, decrease the input sensitivity from ~1V to ~.6v)? My guess would be to decrease the grid resistors on the kt88s to about 600R. The reason for the question is that my source is a modded squeezebox and the output level is insufficient for some older, quieter recordings.

[JoshK]

Glad you got it working reasonably.  I was going to chime in, but wasn't sure what to tell you without doing some research myself.  It sounds like the input tubes aren't matched very well so you need the difference in the cathode bias to make up for the difference in gm of the tubes. 

Re, your other questions....  increasing gain isn't without trade-offs.  It basically redesigns the amp.  If you have a spectral analyzer you can test whether the distortion from reducing feedback (will increase gain) rises dramactically.  But basically you can reduce feedback and thereby increase gain.  You'll need to do some math to determine how much feedback you have now and how much you need to reduce it to get required level.  Your ears might be enough to tell you if you like the results.  The good thing is it should be easily reverseable if it turns out crap.

[gitarretyp]

Glad you got it working reasonably.  I was going to chime in, but wasn't sure what to tell you without doing some research myself.  It sounds like the input tubes aren't matched very well so you need the difference in the cathode bias to make up for the difference in gm of the tubes. 

Re, your other questions....  increasing gain isn't without trade-offs.  It basically redesigns the amp.  If you have a spectral analyzer you can test whether the distortion from reducing feedback (will increase gain) rises dramactically.  But basically you can reduce feedback and thereby increase gain.  You'll need to do some math to determine how much feedback you have now and how much you need to reduce it to get required level.  Your ears might be enough to tell you if you like the results.  The good thing is it should be easily reverseable if it turns out crap.

I did some rough calculations of what needs to be changed in the feedback loop to obtain the gain i need. My only question now is: do i risk damaging anything by altering the feedback loop? I don't have a spectral analyzer, but i do have an oscilloscope. What should i look/listen for when reducing the feedback?

PS: What defines the amount of feedback? Say i have 20db closed loop gain and 40db open loop. Is the amount of feedback then 20db?

[Audiopro]

You won't damage anything by reducing the feedback. Use your scope to make sure the amp is stable, and use your ear to tell if you like it. You can always go back. One other thing I see is the feedback loop is connected to the 8 ohm tap. Typically it goes to the 16 ohm tap. This will affect the gain as well, and will affect sound quality-another easily reversable tweak to try.

[gitarretyp]

You won't damage anything by reducing the feedback. Use your scope to make sure the amp is stable, and use your ear to tell if you like it. You can always go back. One other thing I see is the feedback loop is connected to the 8 ohm tap. Typically it goes to the 16 ohm tap. This will affect the gain as well, and will affect sound quality-another easily reversable tweak to try.

Interesting point regarding the 16 ohm tap. There is not a 16 ohm binding post on the amp but there may be a 16 ohm tap on the OPT that was tied off. Is it convention to use the 16 ohm tap or is there an advantage?

[JoshK]

PS: What defines the amount of feedback? Say i have 20db closed loop gain and 40db open loop. Is the amount of feedback then 20db?

Looks right to me.