D-LA2A Support Thread

[ruffrecords]

This is very odd. Does it work OK with a regular 100K pot?

Is it a PCB or point to point build?

Cheers

Ian

 

[Zander]

This is very odd. Does it work OK with a regular 100K pot?

Is it a PCB or point to point build?

Cheers

Ian

Ian thanks for taking a bit of time to help. You can see pictures of the unit on page 81 of this thread (a few posts back). Yes everything works just fine with a pot but I rewired the unit. I'll put back in a bot just to be sure.


Zander

 

[Zander]

Ian thanks for taking a bit of time to help. You can see pictures of the unit on page 81 of this thread (a few posts back). Yes everything works just fine with a pot but I rewired the unit. I'll put back in a bot just to be sure.

Zander

Yes, works identical as before. Too bad. Should I try to move the 75K to the wiper end?

 

[ruffrecords]

Ian thanks for taking a bit of time to help. You can see pictures of the unit on page 81 of this thread (a few posts back). Yes everything works just fine with a pot but I rewired the unit. I'll put back in a bot just to be sure.

Zander

Yes, works identical as before. Too bad. Should I try to move the 75K to the wiper end?

Well that depends on exactly what you are trying to do. Are you trying to emulate a 100K log pot or do you want some sepcific law - the lists of switch positions sort of imply you are trying for 0.5dB steps.

Cheers

Ian

 

[Zander]

Well that depends on exactly what you are trying to do. Are you trying to emulate a 100K log pot or do you want some sepcific law - the lists of switch positions sort of imply you are trying for 0.5dB steps.

Cheers

Ian

I want the gain range to be smaller. The .5 increments are there for easy recall and they don't need to be scaled in dB. I don't use the compression at the same time when I need the exact recall. I mainly pass signal through it for the tone it has.

 

[ruffrecords]

Well that depends on exactly what you are trying to do. Are you trying to emulate a 100K log pot or do you want some sepcific law - the lists of switch positions sort of imply you are trying for 0.5dB steps.

Cheers

Ian

I want the gain range to be smaller. The .5 increments are there for easy recall and they don't need to be scaled in dB. I don't use the compression at the same time when I need the exact recall. I mainly pass signal through it for the tone it has.

OK, so the exact increments do not matter so much as long as they are repeatable. And you want it to be similar to the 100K log pot but with recall - right?

In which case how about roughly 2dB steps from 0dB attenuation to just over 40dB. WOuld that do? Remind me how many positions you have stopped set the switch for.

Cheers

ian

 

[Zander]

OK, so the exact increments do not matter so much as long as they are repeatable. And you want it to be similar to the 100K log pot but with recall - right?

In which case how about roughly 2dB steps from 0dB attenuation to just over 40dB. WOuld that do? Remind me how many positions you have stopped set the switch for.

Cheers

ian

No, I need a smaller gain range, about a quarter would do: Vout/Vin = 1/4 <=> Vout = 20K/82K * Vin at full gain.
I changed the 5K and 75K resistor for 82K. With RP 82K and the switch about 20K depicted in scenario 2.
I have the front panel engraved for 21 positions.

Thanks again Ian!

 

[[silent:arts]]

Scenario 2 is the one you want, but with -V = GND and potentiometers pin 1 and 3 exchanged.

 

[Zander]

Yes, I didn't state that but I wired it as you say though.

 

[ruffrecords]

OK, so the exact increments do not matter so much as long as they are repeatable. And you want it to be similar to the 100K log pot but with recall - right?

In which case how about roughly 2dB steps from 0dB attenuation to just over 40dB. WOuld that do? Remind me how many positions you have stopped set the switch for.

Cheers

ian

No, I need a smaller gain range, about a quarter would do: Vout/Vin = 1/4 <=> Vout = 20K/82K * Vin at full gain.
I changed the 5K and 75K resistor for 82K. With RP 82K and the switch about 20K depicted in scenario 2.
I have the front panel engraved for 21 positions.

Thanks again Ian!

OK, we need to be careful here to differentiate between attenuation and gain. The reason is the LA2A has quite a lot of gain after the gain pot - I am not sure exactly how much but I think it is around 30dB. So if you want the gain to vary from 0 to 10dB for example, then your attenuator need to cover the range 20dB to 30dB of attenuation.

Cheers

Ian

 

[Zander]

OK, we need to be careful here to differentiate between attenuation and gain. The reason is the LA2A has quite a lot of gain after the gain pot - I am not sure exactly how much but I think it is around 30dB. So if you want the gain to vary from 0 to 10dB for example, then your attenuator need to cover the range 20dB to 30dB of attenuation.

Cheers

Ian

I see what you're saying. It's definitely possible because I have the range available from "1.0": -1.2dBV to "3.5": 11.8dBV, which is about 13dBU or so.
There are other people who have build this unit with stepped controls,  I wonder if they got it working in a smaller range? 
EDIT: I could experiment with a 2K5 log pot soldered to  1K at pos 1 and 100K at pos 3?

 

[ruffrecords]

OK, we need to be careful here to differentiate between attenuation and gain. The reason is the LA2A has quite a lot of gain after the gain pot - I am not sure exactly how much but I think it is around 30dB. So if you want the gain to vary from 0 to 10dB for example, then your attenuator need to cover the range 20dB to 30dB of attenuation.

Cheers

Ian

I see what you're saying. It's definitely possible because I have the range available from "1.0": -1.2dBV to "3.5": 11.8dBV, which is about 13dBU or so.
There are other people who have build this unit with stepped controls,  I wonder if they got it working in a smaller range? 
EDIT: I could experiment with a 2K5 log pot soldered to  1K at pos 1 and 100K at pos 3?

At the moment you have 75K at the top so you have a minimum loss of 12dB. If the gain is 30dB then the overall gain at that position is 30 -12 = 18dB. It might be easier to turn the 100K pot into a 75K resistor in series with a 25K pot. This gives you  a minimum attanuation of 12 dB. The switch then needs to be changed to a 25K pot with whatever steps you want. At least then you are operating in the range you want.

Edit: What is the gain from in to out with the switch set to minimum attenuation?

Cheers

Ian

 

[Zander]

At the moment you have 75K at the top so you have a minimum loss of 12dB. If the gain is 30dB then the overall gain at that position is 30 -12 = 18dB. It might be easier to turn the 100K pot into a 75K resistor in series with a 25K pot. This gives you  a minimum attanuation of 12 dB. The switch then needs to be changed to a 25K pot with whatever steps you want. At least then you are operating in the range you want.

[font=verdana, arial, helvetica, sans-serif]This is what I have now Ian. That's the problem... Only I bought resistors for 23 steps and I shuffled some values, and I came up with a 20K log curve. Its now wired as 82K/20K (resistor/switch). I have the other switch still wired as I drew it a few posts back.
[/font]

Edit: What is the gain from in to out with the switch set to minimum attenuation?

You mean the potentiometer set to maximum or the switch as I drew it af few posts back?

 

[Zander]

Edit: What is the gain from in to out with the switch set to minimum attenuation?

I measured:
Input: -42.21 dBV; -40 dBU; 7.75 mV
Output CCW: -57.60 dBV; -59.82 dBU; 1.31 mV
Output CW: 23.83 dBV; 21.61 dBU; 15.54 V 
When I monitored the sound there was a point where it was louder before CW but measured lower. Here I had my soundcard connected to monitor. For the measurements I had it disconnected. Please confirm if this is a suitable method

Zander

 

[ruffrecords]

Edit: What is the gain from in to out with the switch set to minimum attenuation?

I measured:
Input: -42.21 dBV; -40 dBU; 7.75 mV
Output CCW: -57.60 dBV; -59.82 dBU; 1.31 mV
Output CW: 23.83 dBV; 21.61 dBU; 15.54 V 
When I monitored the sound there was a point where it was louder before CW but measured lower. Here I had my soundcard connected to monitor. For the measurements I had it disconnected. Please confirm if this is a suitable method

Zander

Crikey! -40dBu input giving a +21dBu output implies the gain is 61dB!!! According to the LA2A manual the maximum gain is 40dB so you seem to have an extra 20dB from somewhere. I think the input transformer is 1:1 so there is no gain there which means all the gain must be in the output amplifier. This has some negative feedback which is what sets its gain. According to my calculator, this should be close to 40dB. I suggest you check the negative feedback component values.

Cheers

ian

 

[Zander]

Crikey! -40dBu input giving a +21dBu output implies the gain is 61dB!!! According to the LA2A manual the maximum gain is 40dB so you seem to have an extra 20dB from somewhere. I think the input transformer is 1:1 so there is no gain there which means all the gain must be in the output amplifier. This has some negative feedback which is what sets its gain. According to my calculator, this should be close to 40dB. I suggest you check the negative feedback component values.

Cheers

ian

Yeah no wonder I couldn't scale that properly. It sounds great though... Maybe Volker could tell me where to look?

 

[[silent:arts]]

Crikey! -40dBu input giving a +21dBu output implies the gain is 61dB!!! According to the LA2A manual the maximum gain is 40dB so you seem to have an extra 20dB from somewhere. I think the input transformer is 1:1 so there is no gain there which means all the gain must be in the output amplifier. This has some negative feedback which is what sets its gain. According to my calculator, this should be close to 40dB. I suggest you check the negative feedback component values.

No. Looks like Zander is using the Sowter 4148 wich is 1:4.
The 8940 Output is 3:1.

Just measured my unit: with the 75K resistor and 25K Potentiometer i have a maximum gain of  27dB.

 

[ruffrecords]

Crikey! -40dBu input giving a +21dBu output implies the gain is 61dB!!! According to the LA2A manual the maximum gain is 40dB so you seem to have an extra 20dB from somewhere. I think the input transformer is 1:1 so there is no gain there which means all the gain must be in the output amplifier. This has some negative feedback which is what sets its gain. According to my calculator, this should be close to 40dB. I suggest you check the negative feedback component values.

No. Looks like Zander is using the Sowter 4148 wich is 1:4.
The 8940 Output is 3:1.

Just measured my unit: with the 75K resistor and 25K Potentiometer i have a maximum gain of  27dB.

The 25K/75K gives you 12dB loss so the available  gain  is 27 + 12 = 39dB so close to what I said. However, Zander meausres over 60dB gain so clearly something is wrong,

Cheers

ian

 

[[silent:arts]]

Sure, but since the 1st problem was exchanging a potentiometer with a rotary switch (and different gain range) we should look / compare for this first.
What about 1st channel original with a 100k potentiometer, 2nd channel with the switch?

One other mystery is the 21.61 dBU output which for sure wouldn't be any usable signal / waveform at all anymore.

 

[Zander]

Sure, but since the 1st problem was exchanging a potentiometer with a rotary switch (and different gain range) we should look / compare for this first.
What about 1st channel original with a 100k potentiometer, 2nd channel with the switch?

One other mystery is the 21.61 dBU output which for sure wouldn't be any usable signal / waveform at all anymore.

What do you want me to measure? I did measurements on both already? I measured with a 1K sine wave and measured with a DMM.