[BUILD] New FET/RACK Official Help Thread - Please read first post!

[jdier]

Q13 is crazy off.  You may want to replace it.

OK, ordered and installed new 2N3708 for Q13.  Readings about the same.

 

[Echo North]

OK, ordered and installed new 2N3708 for Q13.  Readings about the same.

Check everything in the meter section including solder joints and values.  Check for continuity between traces.

Mike

 

[cyberlogic]

I doubt it would help to hear it.

With a 1K/0dB tone at the input, set your unit to:

in and out at 12 o'clock
attack in GR OFF
release FCW
20:1
Meter: GR

What is your AC V at TP1, TP15, TP17 and out + with respect to ground (chassis).

Mike

What should these be reading for a Rev A unit?

On both recent builds (A and B), I get (All VAC, ref to chassis GND):-
TP1: A-.140V, B-.145V
TP15: A- 3.72V, B- 3.72V
TP17: A- .67V, B- .71V

Output + to -: A- 10.23V, B- 10.90V

But the output is clipped, I have to reduce the input to around 34 to get rid of clipping and drop out to around 7.30V (+20dBu). Seems like the input gain is too high?

Thanks

 

[Echo North]

What should these be reading for a Rev A unit?

On both recent builds (A and B), I get (All VAC, ref to chassis GND):-
TP1: A-.140V, B-.145V
TP15: A- 3.72V, B- 3.72V
TP17: A- .67V, B- .71V

Output + to -: A- 10.23V, B- 10.90V

But the output is clipped, I have to reduce the input to around 34 to get rid of clipping and drop out to around 7.30V (+20dBu). Seems like the input gain is too high?

Thanks

That's about right what you're getting.

The input pot (t-pad attenuator) on the original units had a very odd taper. It’s not a taper you commonly see in potentiometers today. Our input attenuator has a more common taper, which leads it to be a little “hotter” earlier in the clockwise pot turn.

Mike

 

[cyberlogic]

OK, good to know the numbers are close.

With a 250mV input signal at 1KHz and the output at +4dBu, THD is around .23%. Pretty much the same at 10kHz. Increasing output doesn't add more distortion. The input seems to be the stage that is sensitive to over-driving.

With the taper on the new pot and the fact that this is typically used on line level signals, would reducing input gain a bit not be an option worth considering? I also see a lot of signal bounce when making level adjustments at the lower levels...

Thanks

 

[Echo North]

OK, good to know the numbers are close.

With a 250mV input signal at 1KHz and the output at +4dBu, THD is around .23%. Pretty much the same at 10kHz. Increasing output doesn't add more distortion. The input seems to be the stage that is sensitive to over-driving.

With the taper on the new pot and the fact that this is typically used on line level signals, would reducing input gain a bit not be an option worth considering? I also see a lot of signal bounce when making level adjustments at the lower levels...

Thanks

Certainly, it's on my "to do" list.

The D doesn't see as much of an issue because of the decreased gain. Generally, we just lower the signal coming in a little to compensate but I do plan on documenting a mode as you're suggesting.

Mike

 

[cyberlogic]

What should these be reading for a Rev A unit?

On both recent builds (A and B), I get (All VAC, ref to chassis GND):-
TP1: A-.140V, B-.145V
TP15: A- 3.72V, B- 3.72V
TP17: A- .67V, B- .71V

Output + to -: A- 10.23V, B- 10.90V

But the output is clipped, I have to reduce the input to around 34 to get rid of clipping and drop out to around 7.30V (+20dBu). Seems like the input gain is too high?

Thanks

Also, it seems to me this would affect your recommended Q Bis procedure. With .775VA in, Input at 24, etc, there is significant clipping in the input stage, typically on the positive side of the waveform here.  Shouldn't this be done using a linear signal?

Thanks

 

[Echo North]

Also, it seems to me this would affect your recommended Q Bis procedure. With .775VA in, Input at 24, etc, there is significant clipping in the input stage, typically on the positive side of the waveform here.  Shouldn't this be done using a linear signal?

Thanks

Not with the way the GR FET is configured and not at that level.

 

[Echo North]

I should also mention that this asymmetrical distortion is part of what makes the Rev A so desirable for some. This aggressiveness is why so many changes were made to the Rev B after the Rev A and Rev AB of which only about a couple hundred were ever made.

The addition of the source resistor to Q1, switching the gain stages from FET to all BJT's and lowering the gain were all done to tame the aggressive sound (clipping) of the A and AB.

The nature of this Rev A, A/B beast is to clip and sound aggressive.  That's why it's sought after for things like snare and electric distorted guitars. If you changed the circuit to remove the early clipping you'd just be making something closer to a Rev D. That difference is why we differentiate between the A and D. Then of course the F and G go further to clean the signal.

Mike

 

[Echo North]

One last thing before I quit for the weekend.

Ultimately it's just gain staging. I was just running a signal in at 0dBu and could turn the input past noon before clipping. I'm not sure if +4 was established as standard in 1967, when this circuit was designed, but the original manual uses 0dBu for calibration and you may be much happier operating around that level.

Mike

 

[cyberlogic]

Thanks Mike, I appreciate all the information and you taking the time to consider all this and write back to me.

I really just want to make sure it is working as it would be back in the day, warts and all.

ds

 

[Echo North]

Thanks Mike, I appreciate all the information and you taking the time to consider all this and write back to me.

I really just want to make sure it is working as it would be back in the day, warts and all.

ds

Definitely, anytime.

With these guys you definitely feel like you're trying to hit a moving target because they often changed suppliers of caps, transistors, and transformers.  Back in those days they just popped in whatever they got a good deal on I imagine. We've tried to stay true to the circuit and components using all the info we could get and comparing to originals.

Sounds like your unit is working!

Mike

 

[cyberlogic]

Another question:

How much should the release and ratio controls affect the signal when the gain reduction is OFF (attack pot switch)?

Thx

 

[Echo North]

Another question:

How much should the release and ratio controls affect the signal when the gain reduction is OFF (attack pot switch)?

Thx

As you switch ratios you'll hear a small change as the circuit re-stabalizes.  About 0.5s. But it should settle back to the same level.

Rotating the release will probably change the level by about 0.1dB and maybe a slight change in THD+N. Minimal.

Mike

 

[ganymedestudios]

I finished the assembly of the kit and first tested the ground as instructed and it came back at 0.2 Ω so I moved onto switching the Meter to GR to test the Cathode and -10V diode, neither of them came out to the proper voltage. The cathode that was supposed to be +30V is only +16V and the -10V diode is -6V. Any ideas what the problem could be?

Side note, earlier in the build, I broke a trace and fixed it by soldering some copper wire to the terminals that the trace connected. (Not sure if this could be causing the problem).

 

[Echo North]

I finished the assembly of the kit and first tested the ground as instructed and it came back at 0.2 Ω so I moved onto switching the Meter to GR to test the Cathode and -10V diode, neither of them came out to the proper voltage. The cathode that was supposed to be +30V is only +16V and the -10V diode is -6V. Any ideas what the problem could be?

Side note, earlier in the build, I broke a trace and fixed it by soldering some copper wire to the terminals that the trace connected. (Not sure if this could be causing the problem).

Make sure the broken trace didn't have more connections coming off it. They are not always point A to point B. Sometimes they are Point A to point B, C, and D. Check both side of the board and check the schematic to confirm.

With the unit ON, what is your AC V between ground and each orange secondary wire?

Mike

 

[ganymedestudios]

Make sure the broken trace didn't have more connections coming off it. They are not always point A to point B. Sometimes they are Point A to point B, C, and D. Check both side of the board and check the schematic to confirm.

With the unit ON, what is your AC V between ground and each orange secondary wire?

Mike

The voltage between the ground and orange secondary wire seems to be -0.001V. The trace that I replaced was from R26 to C12. It doesn't appear to have any other connections but I'll keep looking.

 

[Echo North]

The voltage between the ground and orange secondary wire seems to be -0.001V. The trace that I replaced was from R26 to C12. It doesn't appear to have any other connections but I'll keep looking.

There are two orange wires on the secondary.  They should each have their own voltage. Are you measuring AC?

 

[ganymedestudios]

Sorry I must've been measuring the wrong wires, if you're talking about the orange wires running from the transformer, the measurements are 6.3V and 11V And no, I've been measuring in AC.

 

[Echo North]

Sorry I must've been measuring the wrong wires, if you're talking about the orange wires running from the transformer, the measurements are 6.3V and 11V And no, I've been measuring in AC.

Are you able to safely measure those two orange wires with the secondary disconnected from the main PCB. Like can you get your probe into the contacts where it goes into the white plastic header?