TG1 - advice from any one who has built it...

[jplebre]

Hello, can someone raise their hands if they got a rack version 1.2 (fester unlimited) working just by populating boards and doing the hold fix?
Of the lack of starting looking at traces on PCBs, I tried everything but the signal is dead quiet and can't for the life of me figure this thing out

 

[warpie]

So, the last few days I've been reading about different types of amplifiers and by checking the schematic I can't figure what class this circuit is 

Looks like Class A but is it really?

Any idea? 

 

[etheory]

I've just completed an LTSpice simulation of the entire circuit, as I prepare to build mine at some point soon (crosses fingers).
The simulation works perfectly (as I guess you'd expect).

The process and discoveries I've made during simulation I am hoping will help me to choose the right AOT resistors (the descriptions I've read and found so far are OK, but they don't truly help me understand what the circuit it really doing).
However I'll share anything I find and write a very thorough description of exactly how to choose the AOT resistors, a detailed calibration procedure and an even more comprehensive description of how the circuit works than I've seen so far.

What I didn't realize before was that essentially the main gain element is a long-tailed npn pair with a pair of zeners in each of their collectors.
The tail current is varied, which changes the current through the zeners, hence changing their resistance, hence changing the gain of the stage.

In a way this shares a lot topologically with something like a diode-based VCF like the EMS Style VCF's - which I wasn't expecting at ALL.
I think the way the original schematic is drawn partially obscures this fact, but if you lay it out slightly differently, it's clearer that this is how it works.

Very cool.

Anyone have a preference for the 600:600 output transformer to use? I was looking at edcor/carnhill but if there is something that really shines here I'd love to know!

 

[etheory]

Crazily enough, the simulation not only works and compresses, but reveals a LOT about the circuit operation.
I'll write up something soon once I "calibrate the simulation".

I am still surprised at how similar the circuit is to a synthesizer VCA.
I expected some crazy magical circuit design, but it really is pretty simple stuff!

I think replacing most of the AOT resistors for trimmers is probably a really good option too.

 

[haima]

This is VERY cool etheory! Keep us posted

 

[etheory]

Looks like Class A but is it really?
Any idea? 

I take it you are specifically talking about the output amplifier correct?

It is "Class A" (whatever that is) in my definition if the transistors don't turn off throughout a full cycle of a sinewave.
I mean, it gets funky when you talk A/B cause things are kind of off and kind of on, yadda yadda.

By my loose definition, yes, this is Class A, up until it outputs around +-9V into 600ohms at which point the waveform initally clips asymmetrically, and then rather abruptly top and bottom, at which point the output transistors start looking more like they are conducting in Class B for the tiniest portion where they share the waveform duties, then the entire thing falls into a screaming clipping heap around +-11V. Not that outputting those sort of voltages before clipping is anything to scoff at. It's a pretty cool amplifier design, and seems to perform very very nicely on the simulator.

This is all very hand-wavy, but, long story short, it kinda sorta basically is Class A, yes.

 

[bruno2000]

I've just completed an LTSpice simulation of the entire circuit, as I prepare to build mine at some point soon (crosses fingers).
The simulation works perfectly (as I guess you'd expect).

The process and discoveries I've made during simulation I am hoping will help me to choose the right AOT resistors (the descriptions I've read and found so far are OK, but they don't truly help me understand what the circuit it really doing).
However I'll share anything I find and write a very thorough description of exactly how to choose the AOT resistors, a detailed calibration procedure and an even more comprehensive description of how the circuit works than I've seen so far.

What I didn't realize before was that essentially the main gain element is a long-tailed npn pair with a pair of zeners in each of their collectors.
The tail current is varied, which changes the current through the zeners, hence changing their resistance, hence changing the gain of the stage.

In a way this shares a lot topologically with something like a diode-based VCF like the EMS Style VCF's - which I wasn't expecting at ALL.
I think the way the original schematic is drawn partially obscures this fact, but if you lay it out slightly differently, it's clearer that this is how it works.

Very cool.

Anyone have a preference for the 600:600 output transformer to use? I was looking at edcor/carnhill but if there is something that really shines here I'd love to know!

Thanks for doing all of this work.  It's going to inspire me to dust off my boards, heat up the iron, and put some together.
Best,
Bruno2000

 

[Biasrocks]

Anyone have a preference for the 600:600 output transformer to use? I was looking at edcor/carnhill but if there is something that really shines here I'd love to know!

I ended up using a 1166 type transformer with a step up, works a treat.

Hairball has one.

http://www.hairballaudio.com/shop/product_info.php?cPath=24&products_id=63

Regards,
Mark

 

[late]

What is standard value for AOT7?

In Igor's docs it is 3.3k but SLB said that it is 33k.

I get +27,5 and - 27,6 from my psu, is it ok? Or do i need get it closer 28?

 

[late]

One more thing. Can i calibrate vu meter that way?

Send 1khz 0.775v to TG1 and switch it to THD mode. Then trim with 100k trim that the output is same.

Then trim vu meter with 22R and 200R trim that it shows "0". Like in this picture.
http://i50.tinypic.com/wv4vu8.jpg

Does this works?

 

[etheory]

What is standard value for AOT7?

Forgive me if you know this already, but each AOT resistor specified in the circuit would have a different "standard" or, better put, "starting" value.
I found some info in Igor's docs about what he generally uses for these, but it's supposed to be based on measurement.

I get +27,5 and - 27,6 from my psu, is it ok? Or do i need get it closer 28?

Those values are within 2% of what they should be. That's pretty close if you ask me, I'd say you are fine.

 

[jplebre]

etheory, maybe you can help me.
I have a bit of a signal drop. unit works almost as it should but signal is down by about 20dB. I don't seem to have the make up after the gain reduction transistor pair.

VT8, VT10 have signal on their bases, but the collector of VT9 only see's noise on scope.

What should the voltages be around this area?

Or maybe if you have any other idea why there's such a big drop in signal I'm all ears

 

[etheory]

etheory, maybe you can help me.
I have a bit of a signal drop. unit works almost as it should but signal is down by about 20dB. I don't seem to have the make up after the gain reduction transistor pair.

VT8, VT10 have signal on their bases, but the collector of VT9 only see's noise on scope.

What should the voltages be around this area?

Or maybe if you have any other idea why there's such a big drop in signal I'm all ears

VT9 is basically a current source for VT8 and VT10, so it will have a pretty small signal as a voltage, but be providing around 1.6mA of current.
Check the voltage across R27, the 12K resistor, is it around 20V?
Check that R23 and R26 are 3.3K
Check that R22 and R25 are 7.5K
You should see a small audio signal across R22 and R25 and about 6V of difference across them.

 

[late]

I have weird thing. If send audio to channel 1, same signal comes out from channel 2 a little bit. I thing it come by ground some how. What is wrong?

 

[jplebre]

I have weird thing. If send audio to channel 1, same signal comes out from channel 2 a little bit. I thing it come by ground some how. What is wrong?

Are you sharing the main board power?
you should be linking pSU -> main board -> input board
not PSU-> main board -> main board

play with trafo placement. In some units it's enough for crosstalk not sure on this one.

picture of the beast's guts?

 

[late]

Yes sharing. Now i broke it a little more, shieetjh!! Clean mode wont work anymore. ROP5 burned and maybe relays. Will send pics after i get it fixed.

 

[late]

Working again. But the second channel compressing/distort way more. Here is the clip. Hold is 10 and attack/release fastest possible.
https://dl.dropboxusercontent.com/u/20579799/TG1.mp3

 

[late]

Noticed that popping stops when hpf is switched to 40Hz or so. But can't understand what doing this.

 

[thomasdf]

I'd like to dive into the building of one of those lovely comp, as I am looking for something colored to put across my mixbus with gentle compression (4dB max). I need to start reading this thread carefully
Can someone give an approx overall cost of the build?

Many thanks in advance

Thomas

 

[bieckmusic]

Hi Thomas,

I love my T G clone!  I don't have transformers installed yet just the 1:1 ratio with IC's per uncle festers notes. 

I purchased the vintage transistor set from Electrochronic, he was very helpful as it's not the easiest build. 

All in I've got about a $1000 into the project, with mirrored vintage style meters, etc...  $80 for a pair of stepped output switches.  My PCBS didn't come with a power supply so i used a Sontec power supply which seems to work fine.