DIY D1 I/V STAGE

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Input voltage swing

Hi!

I'm re-posting my question here as it looks this thread is more actual then Promitheus's one.

I just finished my take on a D1 I/V based on the original? (post #1) schematics but didn't connected to my AD1865 yet as I'm not sure it is all right as it is.

I noticed two things while adjusting the input voltages for 0V:

1. The input is slowly rising to around +50mV as it warms up. After 10mins T2 warms up to ~50C while T1 stays ~30-40C, so wouldn't normally need a heat-sink but I can mount one. I expect that with a heat-sink it will settle ~10-20mV.
2. On powering up both inputs are at about -3.2V, quickly go down to 1V then 100mV in about 10sec, to 50mV in 30sec and as T1 is warming up it reaches 0V in 2-3min and very slowly rising after. Without heat-sink it probably goes up to +50mV.

- Is this the expected behavior of how this circuit should work?
- Wouldn't that startup voltage swing damage the DAC chip?
- 50mV DC on input is within limits or should I add a heat-sink to keep it lower?

Btw. I'm using Salas shunt reg with 4 x LM329 as voltage references to have stable output.

Many thanks,
Zsolt
 
with that much voltage I would put some zeners at the input before attaching the dac. I presume you are running caps on the output anyway? all these type of things vary with the mosfet used, but you are right to zero the input DC after its warmed up.

have you measured how much the shunts vary on start up? evem with lm329 it wont be all that stable for the very same reasons you are getting some change in the IV. mosfet and resistor temp-co

I would run a heatsink whether the fets overheat or not
 
Last edited:
Yes I have output caps. I'm using IRF610s.
With LM329 the shunts vary under 5-10mV over 20mins, IMO this is very good for 29V output.
OK, I will fit a heatsink to be on the safe side.

The smallest zenner I can found around is 2.4V... not too different. Maybe a simple 1N4148 diode (or two) would be fine for a +- clipper.
 
Last edited:
With LM329 the shunts vary under 5-10mV over 20mins
I mistaken with another measurement... with 4 x LM329 my 28.9V output Salas shunt varied ~ 80-90mV over 5mins - until it warmed up to a stable temperature.

A small heatsink on one side already managed to cut down warming up variation to ~10mV.
I'm still unsure about the -3.2V startup negative offset so no music yet :(
 
Last edited:
As others noticed similar behavior at startup and small drift around 0V with temperature change it seems that this is inherent to the design.

After lowering a bit the negative side CCS draw my ~12K/W heatsink stayed nice ~50-55C even after several hours. The I/V from the 10th minute on stabilized at 5-10mV and remained there. Even opening the case top offset drifted with -10mV so I'm curious how would it be on summer.

So I connected my AD1865 DAC and listened several hours to all sorts of music.
I'm yet undecided if I like this or my passive I/V with SE Jfet preamp better. will do some comparisons soon.
All in all it sounds very good :)
 
Member
Joined 2008
Paid Member
I mistaken with another measurement... with 4 x LM329 my 28.9V output Salas shunt varied ~ 80-90mV over 5mins - until it warmed up to a stable temperature.

A small heatsink on one side already managed to cut down warming up variation to ~10mV.
I'm still unsure about the -3.2V startup negative offset so no music yet :(
Did you ever figure out what to do about that -3.2V startup/power down voltage? I'm getting the same thing. I'm getting music, but for testing I waited until the input voltage got down to near zero before connecting the CD player. Does anyone know for sure whether -3.2V on the input will damage the PCM63P?
 
Member
Joined 2008
Paid Member
Well, in the absence of any replies I decided to bail out on the problem and just use a relay and an NE555 timer circuit to delay connecting the I/V and the DAC chip until the voltage settled. This seems to take about three minutes so I set the timer for about 3:40. The formula for setting the timer didn't at all resemble reality in this case (the timer took a lot longer to trip than advertised), so I had to do it by trial and error. I wasn't using the heater windings on the main transformer so I sued those to power the 555 power supply. It all fit on a round circuit board about 1-1.5 in or so in diameter.

Photos of my version of the Pass D1 I/V are below. Believe it or not I did plan this out.
 

Attachments

  • Pass D1 I-V top.jpg
    Pass D1 I-V top.jpg
    249.1 KB · Views: 455
  • Pass D1 I-V bottom.jpg
    Pass D1 I-V bottom.jpg
    275.3 KB · Views: 427
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.