Mods for perpetual technologies dac?

Has anyone attempted to mod the p3a dac from perpetual tecnologies? I understand certain mods are offered for a fee by Modwright. This includes changing the opa134pa for opa627ap and various cap and resistor replacements for more exotic types.
I was wondering if it was worth attempting it on my own and if other members could give any advice .
 
Hello Protos,

I´ve just received my P-3A and had a short look inside. I think it´s not that difficult to modify as most parts can be reached very easy.
What I will do is along the lines of the Modwright mods.

Change the opamps for OPA627
Change the caps for Panasonic FC and maybe change the 1000uF to 2200uF
Change the resistors near the opamps with dales or holcos
Change the standard power supply with a nice 9VAC toroid (65VA)
Change the diodes for some FREDs

I´ll take some pictures when I do the mods.

But first I will listen a bit more, first impressions are not that good but I think (hope) some running in will better this. I´m using it with my Marantz CD10 (LClock, AD845, OPA2107, OPA627, Pan FC etc.) and at the moment it´s only marginally better in a few ways.

I tried a Chord DAC64 a few weeks ago and this really blew the Marantz away in every aspect and I hope the P-3A will do the same.

Has anybody tried the P-1A? Tests are generally very good but I also found some very negative reviews on the web.

william
 

Attachments

  • p-3a 5.jpg
    p-3a 5.jpg
    87.1 KB · Views: 1,863
Hi Wuffwaff
Well in the end I changed to opa 627 and hexfreds.This was a few months ago.I feel the 627 upgrade was really worthwhile since it fleshed out the sound and evened out a bit of the midrange forwardness of the old opamps.More dynamics too.The hexfreds were bit more subtle giving probably better dynamics and timing.However it is hard to say exactly the extent because you can't A/B with your previous set-up can you?I think the 627s is a must upgrade though. I have got a set of standard BGs for the whole of the P-3 and I am probably goig to put them in at some stage although I think they won't make a big difference.I am using the monolithic PS but I guess you can use the much cheaper alternative of a good trafo. I also use the P1a with usually good results .It gives a smoother presentation more "analog" like as they say.However with the mods the P3 is quite strong on its own now and the P1a might not be worth the extra money depending on your point of view.In some cases it could sound a little too smooth if you know what I mean.Depends on the recording and system you have.Sometimes it gives a more rich deeper soundstage while the modded P3 is more upfront and punchy. I know it is hard but maybe you can audition one in your system.
I am using a Teac Vrds-25 superclock/ps and it has been op-amped with AD ? from LC .(I don't know because somebody did it for me when I didn't have the knowledge to do it myself). I think the P3 sounds quite a bit better with respect to almost everything.I don't know if changing the resistors and film caps in the P3 is really worth it.Maybe you should try it (I have the impression you are more experienced than me) and give some feedback.
 
Hello Protos,

thanks for the info. Just came back from a week in Sweden and the DAC has improved over this week so I´ll wait a bit more before I change anything. The sound is quite good now but somehow a bit "slow" if you know what I mean.

After reading your answer I think the first mods will be the opamps and the caps (have them lying around).

I have also ordered a new transformer.

Can you tell me the difference between the standard and upgraded powersupply?

william
 
mods P-3A

just a little progress report,


changed the OPA134 for 627 and changed the two caps for the opamps power supply from muse 470uF to panasonic FC 680uF.

Didn´t touch anything else cause I want to wait until my transformers arrive.

Sound has improved by the opamp change but not that much. It sound a bit better now than the modified CD10 now (especially in the highs it sounds more naturall and room and placement are a bit better).

Bought a second one (488€ incl. 2 years waranty) for my upstairs stereo and I will leave this unmodified to see if all the mods really work.

william
 
The 627s need some breaking in of course, they sound a little bit heavy at first.. I recently changed the caps around the op-amp supply (can;t be sure without schematic -have you found one?) where there are two nichicon muse 220uf green caps (are you sure 470uf?) with two BG FK 220uf and BG standard caps for the rest (47uf mostly) and sound seems to have more authority in bottom end and perhaps more dynamic.The FK's were four times the size of the muse so were a tight fit. I am thinking of changing the main power caps but it is going to be too tight I think with the BG's.
My freds are 8A .
Desoldering plated thru boards is a bit of a hassle for me -can't seem to suck all the solder from the hole so then I have to redrill them.
 
Protos,

you´re right. They were 220/25V and now 470/25V Panasonics.
Yes desoldering is a bit troublesome. I cut all the little feet of the poor OPA134 before getting them out..........

I´m still waiting for the transformer to arrive so nothing new here.

When really modifying the thing I think it would be convenient to take it out of the case and into a bigger one with more space for power supplies etc.

william
 
de-soldering

Hi,

Try adding a few drops of a no-clean liquid flux to the de-soldering wick before use, I find that this helps to suck up almost all the solder from a hole. If you have already tried to de-solder a joint, then wait for it to cool down and re-solder it before trying this.


Thanks
Raj
 
mod p3a

Here is my plan, replace Opa 134 with Opa 627BP,replace some Fred diodes, replace 47uf,100uf,220uf Capacitors with BG Capacitors. Replace digital input resistor(SMD) 110 Ohms with Caddock MK132 resistor and some resistors around the Opa amp. Protos, what cap did you use around the Opa amp 627?
 
Hi, I also had great results with the modification of my P3-A DAC.
The mod that really made a difference was of course the change of the OPA134 to high quality OPA627 I also changed all the diode to HEXFRED ( this one made a huge difference) the 4X 47uF and the 100uF was changed to black gate with great results I also wanted to change the 220uF but the black gate I ordered was just really too big to fit on the board and don't touch anything else. I also changed almost all the little film capacitor (since wima don't make 5600nF I has not able to change those) to WIMA high quality with also really great results! The last tweak I made was the change of the low quality RCA output to high quality Vampire female connected with jena labs wire and silver WBT solder. This is about all the tweak I already made and I am planing to modify this DAC as much as I can!

Teetee: What are you refering to when you say the digital input resistor(SMD)? Are you talking about the surface mount resistor?
If yes what are you doing to disolder them?

One other tweak I know is possible for the P3-A is the bias of the OPA627 ( as far as I know you can bias those OPAMP in class A but I really don't know how)
 
mods p3a

So far, I replaced opamp with 627, replaced 47uf, 100uf, 200uf capacitors with BG. Replaced diodes to FRED diodes.

dismalonyx, it is ok to fit the 200uf BG Cap if you bend them a little away from each other. Where did you get the Wima Cap?
I had difficult time to find some Wima Caps on Internet.
I haven't received my Caddock resistors yet so I did nothing to
the resistors.
For those SMD (surface mount) digital input resistors, just heat them with solder gun and push them away gentley.
Don't know how to bias those Opamp either, leave this to whom did it before.
I still let my moded DAC burn in a bit before do a crital listen.
 
opamp biasing

Hi,

wouldn´t know how to bias an opamp. As far as I know this is done by very small resistors inside the opamp and the only way to change this is to open it and have a look:bigeyes:

The only thing I can imagine when people talk about biasing opamps is that they change the power supply voltage????


william
 
Op-amps in class-A

Frederic,

The easy way to bias an op-amp to class-A is to pur a resistor between the negative voltage input an the signal output.
In the case of the OPA627, it's between pins 4 and 6.
You can use google, you'll find plenty of information on the net.:idea:
 
Perpetual P-3a and Opamp Class A Bias

On an opamp output there are normally two output transistors, one pulling positive and one pulling negative. They normally have a small current flowing through them. To bias them into single ended class A a larger current must flow to either the +ve supply or the -VE supply. If the current were from a resistor from the -ve supply then the transistor which pulls down on the output would switch off and the transistor which pulls up would turn on more than without the resistor fitted, to let say 5mA. This is now a single ended class A output stage.

It unfortunately has problems as it is not linear due to the current variations which would occur through the resistor when the output voltage on the opamp varies with the music. This is basically psu voltage - output voltage divided by the resistor value. There is a dynamic change in current through the resistor as the opamps output changes which the opamp has to compensate for to produce a linear output. This is due to the fact that the dynamic impedance of the resistor is actually the resistors value, ignoring parasitic inductance and capacitance which exist in the real world. If you had a +/-12V supply and wanted a current of 5mA you would fit a 2.4k resistor.

Also there is now a lowish impedance path from the PSU pin to the opamp output which will try and feed psu noise into the opamps output. The opamp will also have to compensate for this to reject this induced noise on its output pin.

There is however a solution - the fitment of constant current source which also sinks 5mA from the output pin. The current would be the same as before with no signal, but would also be the same under dynamic conditions as the source of current is a constant current source and not a resistor. This removes the linearity problem.

The constant current source (CCS) current does in fact change a small amount under dynamic conditions and this can be calculated from actual measurements. 12V is placed across the CCS and its current measured - the supply is then taken to 14V and the current measured again. The current may have gone from say 5mA to 5.002mA giving a change is 2uA for a 2V change in output as compared to the resistor which would have been 5mA to 5.833mA - a change of 16% as opposed to 0.04% for the CCS. The dynamic impedance of the CCS is now calculated as 2V/2uA = 1 million ohms. This means that the grunge from the supply has to effectively go through 1 million (1M) ohms to effect the output - a factor of 416 time better than the resistor. It is also highly linear compared to the resistor - the class A bias only changes by 2uA over the 2V swing as opposed to 0.833mA for the resistor.

The simplest way would be to fit (they are diodes so polarity needs to be observed) a 5mA constant current diode between either the +ve or -ve supply pin to the output pin 6 to bias into class A - giving all the positive improvements of class A bias but very small negative.

A J511 (Siliconix) diode has very similar specs at 4.7mA.

Hope this helps.

David
 
looking for a used P-3A

Hi, i am trying to find a used P-3a on the net, but haven't been sussessfull yet. I tried E-bay, Audiogon, Google but no luck so far. Any suggestions where to look?
Also looking for the complete schematics.
Any help would be highly appreciated.

Thanks,

Arnaud
 
Re: Op-amps in class-A

carlosfm said:
Frederic,

The easy way to bias an op-amp to class-A is to pur a resistor between the negative voltage input an the signal output.
In the case of the OPA627, it's between pins 4 and 6.
You can use google, you'll find plenty of information on the net.:idea:
-------------------------------------------------------------------------------
You also inject noise from the power supply unless you are quite careful. CCS and two fet versions sound much better, in that order.
 
Re: looking for a used P-3A

catzjes said:
I found a used P-3A on www.adverts.hififorsale.com for a good price (€ 470 including shipping from UK to the Netherlands).
There is one more for sale for GBP 300,00 (€ 450,00) so if you are interested, take a look.
Still looking for the schematics though.

Regards,

Arnaud

Congratulations, very reasonable price., I use one too (especially icm with a Beringher DSP). Paid 600 euro a year ago.
(look here)
Are you planning to do some mods (mine is still standard)? opamps, caps, rectifiers powersupply etc. And what about the P1?
Ronald