Deconstructing Rowland 7 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st August 2011, 01:24 PM   #1
diyAudio Member
 
analog_sa's Avatar
 
Join Date: Aug 2002
Location: Sofia
Default Deconstructing Rowland 7

After a mishap with one of the potted modules of my Model 7 i got curious and took a peek inside. Always wondered why it sounded dull and attributed this to filter caps, passive components and execution. After all, based on this very useful post of BrianEno Threshold SA/4e vs SA/1 and the quote from the Model 7 user's manual "Lack of overall or nested negative feedback eliminates time domain distortion" made me think that conceptually at least, this amplifier was indeed excellent.

Output stage is identical to the Threshold Stasis amps where it is similarly not within a NFB loop. John Curl once mentioned being a consultant for the early JRDG amps and this meshed well with the Mk2 input schematic shown in BrianEno's post above. Can you possibly combine Curl and Pass and not achieve greatness?

A bit of historical background may be useful here. The Model 7 was initially produced in non-clandestine fashion. Eventually a Mk3 with an epoxied front end was introduced, followed by a "transimpedance" upgrade module which allowed Mk3 owners to obtain an even better performance. My set of Mk3 amps came with the "transimpedance" module installed.

So, after a bit of digging through some of the hardest epoxy known to man i came up with a circuit diagram. It may not be 100% correct as the reverse engineering exercise was a bit messy but the results are, nevertheless, interesting.

To begin with, there are two medium quality double opamps per channel which serve as balanced to single ended convertor, allow adjustable voltage gain, output dc servo and provide the important low impedance drive for the transimpedance section. I find this quite amusing. A high end manufacturer seems to have more scorn for audiophiles and their hearing than even hard core objectivists. So, the signal has passed through hundreds of opamps, what can another three do?

Adding further insult is the power supply arrangement for the opamps: 400uf per rail followed by a pair of 12v zeners, followed by about a foot of wire. No local decoupling of any sort that i could find. Imagine the poor audiophiles trying to find a matching zero NFb preamp for their Rowlands. The travesty of the high end.

And then there is the actual power amp. Unlike Rowland's claims it is quite the opposite of feedback free. It has, however, a rather interesting circuit. I certainly do not understand it completely and fail to translate it to standard topologies. The output, biasing, servo and input stages are clear but i really get lost with the workings of the current mirrors. If anyone feels up to analysing this ciruit it will really be appreciated. The front end seems a bit similar to the Chinese Model 9 clone circuit which i am not even sure is entirely correct Jeff Rowland model-9 amp schematic,really?

I simmed both Mk2 and Mk3. Unsurprisingly, most of the distortion is generated in the output stage. It seems excessive, so maybe something is not quite right with my sims. The "stasis" pair is certainly the most critical part of the design. I couldn't find models for the 2SA1073/2SC2523 originally used and replaced them with 2SC2922/2SA1216 which should be at least as good. The servo is omitted for simplicity but would normally connect to the grounded side of R36 on Mk3 and R41 on Mk2.

As much as i like the Mk2 curcuit, transistor matching and precise setting of the input current sources is probably not easy. Could it be the reason JRDG quietly abandoned this circuit?

R35/R37 in the transimpedance sim were actually inductors. Forgot to measure them.

Eventually, i will probably build several front ends while retaining the Threshold output stage. Maybe a Threshold, a Mk2 and a Mk3 but with proper opamps, properly powered. Looking at the really high distortion at output i am not so convinced it is a great idea, but converting to a standard emitter follower output stage under a feedback loop may just be too hard for me.

I hope these circuits are of some interest to others as well. Any suggestions, corrections and modifications are most welcome.
Attached Files
File Type: pdf transimp.pdf (25.1 KB, 248 views)
File Type: pdf transimp full output.pdf (25.4 KB, 167 views)
File Type: pdf model7Mk2 full output.pdf (29.0 KB, 183 views)
  Reply With Quote
Old 21st August 2011, 01:32 PM   #2
diyAudio Member
 
analog_sa's Avatar
 
Join Date: Aug 2002
Location: Sofia
And here are the distortion spectra where VAS indicates distortion prior to the output triple in the Mk2 and at the bases of the "stasis" transistors in the transimpedance version.
Attached Images
File Type: jpg Mk2 50vpk distortion.JPG (77.0 KB, 307 views)
File Type: jpg Mk2 50vpk distortion VAS.JPG (74.8 KB, 299 views)
File Type: jpg Mk3 50vpk distortion VAS.JPG (82.9 KB, 297 views)
  Reply With Quote
Old 21st August 2011, 01:37 PM   #3
diyAudio Member
 
analog_sa's Avatar
 
Join Date: Aug 2002
Location: Sofia
Sim files
Attached Files
File Type: txt model7Mk2 full output.txt (24.7 KB, 83 views)
File Type: txt transimp full output.txt (21.4 KB, 50 views)
  Reply With Quote
Old 28th August 2011, 03:51 PM   #4
diyAudio Member
 
Join Date: Jul 2007
Hi Analog_SA,

Really outstanding and painstakingly job you performed.
Questions is if Jeff Rowland really swapped at it's Series III incarnation of the M7 to the encapsulated frontends to hide the implementation of those IC's that are not usually associated with highend sound.
I think collegues like Nelson Pass and John Curl won't react to your findings out of courtesy for a fellow designer although I would be very interested in their professional opinions.
As you already know I can't comprehend your findings but having had the same amp twice (silver and satin black version) and also for three years a slate grey Series II, I found the transimpedance version the best sounding one on my Acoustats. I would recommend this amp to anyone who has a need for power combined with musicality. Together with the Levinson ML-2 and the Threshold SA/1 I found this Model 7 the best amp I had.
Nevertheless your perspective is a valid one and your entitled to criticise/have an opinion about the designchoises of a particular product although what you effectively say is that the frontend isn't that great (bit of an understatement in light of your remarks) and the outputstage is a copy of Thresholds Stasis section, hmm.
I'm curious what others like Jacco or Zen_mod have to say about the frontend of this M7 because really I can't say anything intelligent about it due to lack of knowledge. It does look pretty complicated compared to the frontend of my SA/1 optical bias though.
Please don't be shy!

Last edited by RobertS61; 28th August 2011 at 04:18 PM.
  Reply With Quote
Old 29th August 2011, 03:30 AM   #5
The one and only
 
Nelson Pass's Avatar
 
Join Date: Mar 2001
I really don't see any Curl influence there. Perhaps you should ask him.

  Reply With Quote
Old 29th August 2011, 09:08 AM   #6
diyAudio Member
 
Join Date: Jul 2007
Quote:
Originally Posted by Nelson Pass View Post
I really don't see any Curl influence there. Perhaps you should ask him.

Maybe John Curl contributed to one of the predecessors of the Rowland Model 7 like the WL-500 that was manufactured for Roger West (West Labs, Soundlab) or the first Series M7.
I posted a thread with that question but got no response as I expected because this is a DIY forum and not a place to discuss who did what with whom...
I can imagine that it's a delicate matter because circuitdesigning is the creative process that's (or should be) the key element of any highend audio firm. I can imagine that as a designer you can get inspired or influenced by designs from others but it becomes thin line when your design leans heavily on another mans work and can't be called truly yours therefore.

Having said this I have not a single indication that's this is the case with any of JR's designs.
I'm still interested as Analog_sa the starter of thread, what are the opinions about his reverse-engineering frontend of the M7 and the 'Stasis' outputsections as he describes it.
As said before I fully understand that NP or for that matter JC, won't comment on a fellow audio-designer's circuit.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Jeff Rowland Model 7 Please Help rebank Solid State 2 13th February 2009 09:44 PM
Rowland Research Amplifier Help! ungie Solid State 97 7th December 2008 02:23 AM
Rowland Model 1 upgrade PH104 Solid State 0 2nd December 2008 10:46 PM
pass vs jeff rowland arthur Pass Labs 35 18th October 2006 05:29 PM
Jeff Rowland ben goh Analogue Source 0 14th May 2005 06:04 AM


New To Site? Need Help?

All times are GMT. The time now is 07:57 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2