has anyone tried the PGA2310 or Wolfson chip yet?

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

Rowland is now using PGA2310 in the Synergy preamp in place of the CS3310, for which it substitutes pin for pin, but runs off of a +/-15V supply rather than +/-5V, meaning less risk of overloading the inputs. (Both boards, however, require a +5V digital supply.)

As far as sound quality, I've heard both volume controls in my system, but I cannot say one chip is better than the other stictly on sonics as the new board with the PGA2310 also has a totally different topolgy. Having said that, I would most certainly use the PGA2310 over the CS3310 if only for the better overload protection: with 15V rails, the former can accept 27V peak to peak inputs.

I've had no experience with the Wolfson.

regards, robert
PGA2310 listening test

After this post, you guys got me to thinking. I have the old board from the Synergy II preamp, left over from the upgrade to the IIi, so maybe I could use it to test a PGA2310. I got to studying board to see how it worked, and I came up with a scheme that would require the least amount of fiddling. The results are very good!

First, the way I read the circuit board (some of this stuff is already published; Stereophile 9/99), it uses a Jensen JT10KBD input transformer that steps down the input signal 4:1. The balanced output of this transformer goes to one CS3310 run at +/-5V then onto a pair of opamps (with a DC servo), converting back to a balanced signal and then to a Jensen output transformer. The step down transformer is probably used to avoid overloading the CS3310 since it can only input/output 1.5V lower than each rail, or about 7Vp-p. Interestingly, there is a CS3310 per channel, and each one was used with one input grounded (ie, not fully balanced).

My idea was to replace the CS3310 with a PGA2310. As I studied it I got to thinking, why not use the PGA2310 fully balanced with each half (L/R) used for (+) signal leg and the other for the (-) leg. Also, since I was going to run it at +/-12V (21Vp-p), why bother with the input transformer? Since the output was inherently balanced, why use an output transformer? Moreover, as the PGA2310 already has an output buffer (bipolar and linear to 1.5MHz per spec sheet; maybe current feedback type?), why bother with the output buffer opamps? And finally, since the spec sheets quote a 0.5 mV offset, why have a DC servo?

So I removed essentially everything from the board, replaced the CS3310 with the PGA2310, cut a few traces, jumpered here and there, replaced the 7805 and 7905 with 12V versions, and added a 7805 for the digital voltage source. I replaced the IIi board in my Synergy with this modified II board after listening to the IIi for a few minutes, and then listened to the new unit that is actually only one PGA2310 per channel and nothing else.

The bottom line: it is great. The noise level dropped (why not?, two less transformers and two less op amps). The preamp is absolutely dead quiet. Dynamics are better and the soundstage is superior; in one SACD track I routinely use for testing, the singer's voice seems to now be really hanging out in front of the two speakers, perfectly centered and the depth is even better.

The bottom line is the PGA2310 is VERY good. It basically ends up being the entire preamp. Measured DC offset is 0.4 to 0.9 mV. You do need a source of less than 600 ohms to avoid increasing distortion (per spec sheet); but most sources fulfill this requirement. I do not know the output impedance, nor the current drive--does anybody have any information on this to share?--so if the output buffer is meager, then the PGA2310 could conceivably need another interface (opamp/mosfet source) with different loads or cables. In fairness, the input/output transformers may provide better rejection of RF interference and the like; I, however, have not noticed any problem since the change.

Regards, Robert

PS Sorry, Hugh, I keep getting side-tracked from building your AKSA amp. Soon, I promise.
Nah - I'm the guy who was able to admit that I couldn't hear a difference when I did the test! Brave, huh?

I also admitted that this might have been a defect in my hifi or my hearing. But, since then, I've upgraded my speakers and plan to repeat the tests soon. I try to keep an open mind.

(I had this reply on my webpage for a while, as I hadn't signed up to these forums at the time. It's still in Google's cache if you're interested. My comparison was limited to 5532 vs OPA2134)

I'm sorry if my webpage offended you. To return to the topic, I'm very pleased with the PGA2310. On my modest system, it's completely transparent. I'm almost ready to build the final version - just waiting for a collegue to etch the PCB's for me....


on the contrary...

it was not offensive at all. anyone who would be offended by your comments should really get a life.

sorry if i was sounding a little condescending myself. your webpage is most useful, and i did not mean to disparage your work - it is excellent. you seem to be very far along in your prototyping, which is more than i can say about my own projects. :eek:

i am not even particularly surprised that you preferred 5532 to OPA2134, as a lot of people don't like the sound of the 2134 (too euphonic?). it is good to keep an open mind as you say though, and i'm looking forward to your results with other devices. definitely put the AD825 and OPA8610/8620 on your list, they seem to be a couple of the "hot chips" around here lately. i'm actually about to try a OPA2604 in my SACD player, to be followed by a AD8620 if i get far enough along in my mods.
I got some feedback from TI/BB. The buffer op amp is very similar to the OPA134. Maximum output current is 35 mA with an output impedance of 25 ohms.

Since many people use OPAx134 as output buffers, it would seem than an additional output buffer really isn't necessary with the PGA2310 when run with 12 to 15V rails.

dorkus, I too like the AD8610. I very happy with in my active crossover and phono stage.

And mhennessy, I very much like your site, it is very imformative. Thanks for sharing.

Regards, Robert
Hi dorkus,

Thanks for clarifying that - I wasn't sure what you meant, and I apologise for taking it the wrong way. As you can tell, I'm new here :)

I'll make a note of the opamps you recomend. I'm hoping that my listing skills (which are improving daily with my new speakers) will be up to the task! I'll report back my findings, but this is still some way off...

rljones - that's intersting information from TI/BB. I agree with what you say about the output buffer - I suppose that my reasons for adding one (as given on the webpage) are a bit spurious, but at the time, the PGA2310 was still a very new device...

Thanks to both of you for your positive feedback! When I've got some news on the preamp, I'll let you know...
I forgot to add that the slew rate of the PGA2310 output buffer is 16 volts per microsecond per TI/BB. This is not so bad for a preamp: it should not slew limit with a 12.5 V output until after 200 kHz. Since most amps go full power by 2 V on the input, 2V would give a theoretical slew limit of 1.25 MHz.

These figures are based on the op amp slew calculator at http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/a741p3.html

Regards, Robert
Re: PGA2310 listening test

rljones said:
So I removed essentially everything from the board, replaced the CS3310 with the PGA2310, cut a few traces, jumpered here and there, replaced the 7805 and 7905 with 12V versions, and added a 7805 for the digital voltage source.

I believe the PGA2311 has the same pinout as the CS3310 but it uses only +/- 5V, unlike the PGA 2310. Most of the specs are the same as the PGA2310, with the exception of max. input voltage(obviously) and a few others.
While the PGA2311 is a direct substitute for the CS3310 (discussed earlier in this thread), the PGA2310 runs at greater voltages (+/-15 V max) and can be daisy chained for multi-channel use via one of the pins (see data sheet). As I mentioned in my description, I'm using one PGA2310 per channel in a fully balanced preamp.

Regards, Robert
hi robert,

hmm, so i re-read your mods of the extra Synergy board you had lying around. i am actually building a preamp for my friend right now... we saw the latest Synergy at a dealer and he really liked it, in operation, looks, and sound. so i am building something like that for him. not Son of Dork, more like Bastard of Dork. :p

anyway how do you think your PGA2310-direct version sounds compared to the latest Rowland board version? i am thinking of this:

- balanced and unbalanced inputs
- unbalanced inputs go thru Jensen transformer to convert to balanced
- balanced signal go through Borbely JFET buffers (simple 2-transistor circuit, DC coupled)
- output of JFET buffer feeds PGA2310
- output of PGA2310 feeds another JFET buffer (same topology)

how does that sound to you? it is still DC coupled from input to output. if necessary it may need a DC servo at the output but i would like to avoid it as much as possible, hopefully some circuit tweaking will take care of it. the PGA2310 by itself should have enough gain right? the JFET gain is very close to unity.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.