DIY conversion of plate amp to servo

0 Members and 1 Guest are viewing this topic. Read 1121 times.

tnargs

  • Jr. Member
  • Posts: 51
DIY conversion of plate amp to servo
« on: 12 Jun 2019, 09:59 am »
Hi everyone

I asked this question on diyaudio and got some generic answers, but now that I have discovered this GRR-specific forum, this might be the ideal place to ask.

I am looking at making a DIY of the GRR or Rhythmik type of servo subwoofer.

I see that there are various types of 'special' plate amp offered by those 2 companies for these drive units, that have the necessary servo connections.

But what if I want to use a different plate amp? What is the necessary modification to use it with the above servo sub drivers?

Has anybody done this at home? What's involved?

I don't think I am up to designing the conversion circuit myself. I would need to see an actual circuit, ideally a reverse-engineering of what Rhythmik or GRR actually use in their servo plate amps.

I basically want to be able to copy what they do, and apply it to amps different from the ones they supply.

Thanks
Arg

S Clark

  • Volunteer
  • Posts: 5875
  • measurement? We don't need no stinkin measurement
Re: DIY conversion of plate amp to servo
« Reply #1 on: 12 Jun 2019, 12:37 pm »
Am I understanding correctly?  Are you asking for someone to give you the patented Rythmic design for free?

tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #2 on: 12 Jun 2019, 01:40 pm »
It's a DIY project.

Peter J

  • Full Member
  • Posts: 1230
  • Hmmmm
Re: DIY conversion of plate amp to servo
« Reply #3 on: 12 Jun 2019, 01:43 pm »
It's a DIY project.

How does that change anything?

gab

  • Full Member
  • Posts: 568
Re: DIY conversion of plate amp to servo
« Reply #4 on: 12 Jun 2019, 01:49 pm »
It's a DIY project.

Well then do it yourself and report back on your results.

gab

tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #5 on: 12 Jun 2019, 02:07 pm »
If anyone from GRR directly wants to comment on some IP issue, then fine.

Other people are welcome to comment if it is helpful to my project. But please keep the nasty accusations to yourselves. I have no ill-intentions whatsoever.

Regards,
Arg

Captainhemo

Re: DIY conversion of plate amp to servo
« Reply #6 on: 12 Jun 2019, 02:08 pm »
How does that change anything?

It doesn't  !

Is there a specific reason why you  want  to use  different amps  ?
These drivers and amps  have been  designed to work extremely well together, you  don't  need  to  change  anything.

jay

tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #7 on: 12 Jun 2019, 02:47 pm »
Is there a specific reason why you  want  to use  different amps  ?
These drivers and amps  have been  designed to work extremely well together, you  don't  need  to  change  anything.

jay

Thanks, jay, for asking the right question. My project is in the early concept phase.

My intention would be to utilise 8 of the SW-12-16FR GRR open baffle subwoofer drivers, in 4 OB cabinets/frames, 2 drivers to a cabinet. I want freedom to experiment with sub placement in the room, hence 4 of them.

I assume that, with OB subs, the driver is specifically chosen for high excursion, so I want to make sure that the amps can deliver that. So ideally I want two 800W/4R amps, each driving 4 drivers in parallel. The A370 amp doesn't seem to have the voltage to fully utilise the 16R driver, unless it is unusual and I have wrongly assumed it has the usual voltage rails for a 370W/4R amp.

I asked Rhythmik about the HX800 amp but it turns out to be 2x 400W/4R amps on one plate, so again, not the voltage to drive the sub drivers to full excursion.

This is when I got a bit despondent and started thinking along the lines that led to this thread, to see if I could do something like buy and adapt a pair of UcD700.

If there is a better solution than the DIY adaptation that my OP is about, then I am all ears, because I was not particularly optimistic about pulling it off without glitches in the implementation.

cheers
Arg

Peter J

  • Full Member
  • Posts: 1230
  • Hmmmm
Re: DIY conversion of plate amp to servo
« Reply #8 on: 12 Jun 2019, 02:48 pm »
If anyone from GRR directly wants to comment on some IP issue, then fine.

Other people are welcome to comment if it is helpful to my project. But please keep the nasty accusations to yourselves. I have no ill-intentions whatsoever.

Regards,
Arg


Arg, I rarely wade into a fray like this, but your apparent attitude concerning intellectual property not belonging to you seems arrogant and self-serving. You posted on a public forum, and yet want to limit the responses to those which further your stated agenda.

Look at it this way, f you developed a proprietary process to do something and it was part and parcel of your livelihood, how anxious would you be to give it up to the world, DIY or otherwise?  If that was the original intention of Rythmic, I imagine it would be in the public realm by now, don't you think?


So you know, I'm not part of GRR or Rythmic, but am a purchaser of both.


Come on man, you have to realize that you were going to inflame some folks here by asking for help to "reverse engineer" a product such as this.




Captainhemo

Re: DIY conversion of plate amp to servo
« Reply #9 on: 12 Jun 2019, 03:12 pm »
Thanks, jay, for asking the right question. My project is in the early concept phase.

My intention would be to utilise 8 of the SW-12-16FR GRR open baffle subwoofer drivers, in 4 OB cabinets/frames, 2 drivers to a cabinet. I want freedom to experiment with sub placement in the room, hence 4 of them.

I assume that, with OB subs, the driver is specifically chosen for high excursion, so I want to make sure that the amps can deliver that. So ideally I want two 800W/4R amps, each driving 4 drivers in parallel. The A370 amp doesn't seem to have the voltage to fully utilise the 16R driver, unless it is unusual and I have wrongly assumed it has the usual voltage rails for a 370W/4R amp.

I asked Rhythmik about the HX800 amp but it turns out to be 2x 400W/4R amps on one plate, so again, not the voltage to drive the sub drivers to full excursion.

This is when I got a bit despondent and started thinking along the lines that led to this thread, to see if I could do something like buy and adapt a pair of UcD700.

If there is a better solution than the DIY adaptation that my OP is about, then I am all ears, because I was not particularly optimistic about pulling it off without glitches in the implementation.

cheers
Arg

Well,  you can use a pair of   the  SW12 08FR's in each cabinet driven by an  A370..  You'll want  a separate amp for each cabinet   as yo uwant to keep the servo leads  as short as possible so driving a pair of cabinets with a single amp  and having he flexiblility to move  the cabinets  around   is  not  going to work.
You  could also  use  3 Sw12 16FR's per cab driven  by  the same single  A370..    Either way,  the  above  applies, you'll want  a single amp for  each   cabinet and  keep wiring  as short as possible.

You  won't  have  any issue  with  SPL from   these

jay

richidoo

Re: DIY conversion of plate amp to servo
« Reply #10 on: 12 Jun 2019, 04:00 pm »
OB drivers are high Q. They have less mechanical damping which allows exaggerated cone movement at high excursion, which only happens at LF. It is bass boost built into the driver, but the drawback is loss of detail at all frequencies because the cone is poorly controlled mechanically. Servo restores damping to the system electrically, giving a low Q system with good transient response. The high Q driver allows the servo to control the cone with less power than a low Q driver would require, especially at high excursion.

They will certainly play loud enough for you, especially with 8 drivers in the room. You don't need more voltage gain with 4 ohm loads.

OB subs need space around them to get the full benefit of increased bass clarity. Make sure your room is big enough area to provide the breathing room for 4 subs, and also big enough in volume to require 8 sub drivers.

mlundy57

Re: DIY conversion of plate amp to servo
« Reply #11 on: 12 Jun 2019, 04:10 pm »
What Jay said. Three 16ohm drivers in parallel per A370 amp allows you to use the full power of the amp without any worries of bottoming the drivers. I have two 8ohm drivers on an A370 and while it doesn’t happen very ofter, I have had some bottoming issues when cranked way up.

What both Jay and I have been talking about with three 16ohm drivers is for open baffle use. The 4ohm GR driver is for sealed or ported use. I’m not sure about using the GR 8ohm drivers in a sealed box but Rythmik has drivers designed for two drivers in a single box.

You can use the Rythmik drivers with any amp as a conventional subwoofer driver but it will not have servo control. To use these drivers with servo control you have to use a Rythmik amp. The drivers and amps are designed to work together for servo control and, as others have already mentioned, that servo design is patented.

Rather than trying to DIY a protected design, you would be better served to discuss you issues with Rythmik. You might spur the development of a new amp model that addresses your issues and has the Rythmik servo control.

Mike

Danny Richie

  • Facilitator
  • Posts: 12708
    • http://www.gr-research.com
Re: DIY conversion of plate amp to servo
« Reply #12 on: 12 Jun 2019, 07:32 pm »
If anyone from GRR directly wants to comment on some IP issue, then fine.

Other people are welcome to comment if it is helpful to my project. But please keep the nasty accusations to yourselves. I have no ill-intentions whatsoever.

Regards,
Arg

I forwarded a link to this thread over to Brian at Rythmik Audio. If he feels it necessary to respond then he can.

As for your concerns. The amp and woofer combinations that you see offered are matched. So in each application the amps have the ability to drive the woofers to full X-Max without over driving them.

Also keep in mind that the amps have to be placed right next to or behind the woofers. Adding additional lengths of cables between them causes problems and delays the controls. So if you want four sets of doubles then you'll need four amps.

tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #13 on: 12 Jun 2019, 11:29 pm »
Well,  you can use a pair of   the  SW12 08FR's in each cabinet driven by an  A370.. 

OK I hadn't even noticed the SW-12-08FR driver. It would certainly fix my amp concerns.

But why is its Fs so much higher?

And what does that mean in practice? (In the ultimate end game, at the highest SPLs and lowest frequencies)

Thanks

tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #14 on: 13 Jun 2019, 12:46 am »
I forwarded a link to this thread over to Brian at Rythmik Audio. If he feels it necessary to respond then he can. ....
I enquired on their website a few weeks ago, and had a couple of emails with someone called Brian Ding. Could be the same Brian. He suggested one HX800 running each 4 16R drivers, but when I realised it is 2x 400W/4R modules, I didn't follow up that amp with him, thinking it is not going to drive a 16R unit beyond 100W. I did ask a follow-up about he HX600, (thinking to myself it would have higher voltage rails), but he didn't answer.

....Rather than trying to DIY a protected design, you would be better served to discuss your issues with Rythmik. You might spur the development of a new amp model that addresses your issues and has the Rythmik servo control.

Mike
Well I hadn't thought of that, but it seems a bit speculative, even though ideal. I kept my questions to Rythmik short and businesslike, " ....I am thinking of running 2 amps to 8x SW-12-16FR drivers, which will share 1 amp between 4 drivers. Therefore high power amps are needed to get high excursions from these 16R drivers. ....is there a high-power option coming to replace the (discontinued amps)?" Once that discussion petered out pretty quickly (and I'm not complaining, people are busy), I thought Oh Well, I'm going to have to try and work out a solution myself, but I'm going to need some help. And here we are.  :)

Back to technical. In my wish to get full excursion from the drivers, I have been calculating on the pretty rough-and-ready basis that the 16R, 200W drivers at least need enough amplifier to get them to 200WRMS, which will require balanced 80V DC rails on the PSU. I would be very, very surprised if the A370 can do this. On the other hand the UcD700 has 90V rails and can do it. The Rythmik HX800 would have been ideal if it was a single module, but if it is 2x400W modules, it can't.

Unless the HX800's two internal modules can be bridged??  :scratch: (dubious look) Even if possible, it would probably be limited to a minimum load of 8R --yes?-- and unable to run 4 drivers.

If someone has a better technical analysis of what it takes in amplification to deliver full excursion from the SW-12-16FR, then I am all ears. Because my calculation above is, how shall I put it kindly, indirect.

cheers
Arg

rythmik

  • Industry Participant
  • Posts: 181
    • Rythmik Audio
Re: DIY conversion of plate amp to servo
« Reply #15 on: 13 Jun 2019, 01:43 am »
SW12-16FR has an DCR of 12ohms.  All of our drivers have taken the approach of 3ohm DCR (4ohm nominal, when one averaging the impedance over frequency).  So if one does the calculation, it takes 4 parallel drivers to make it 12ohm/4 = 3ohms. That was our original goal.  In practice, we do find that A370 (class A/B) series amps, when driving 4 drivers, can lead to amplifier overheat. One can use a fan, or not to play as loud to mitigate the problem.  When driving only 2 SW12-16FR, the A370 series amp is fine.  That amp was introduced more than 10 years ago. Starting 5 years ago, we began to offer HX800 with Hypex class D modules and that completely addresses the amp temperature issue. The next issue is how to put all 4 drivers in close proximity so that servo control remains effective. Since HX800 has  two 400WRMS modules with independent servo control, one can use one module to driver 4 with one module and put it in left and another 4 in right in mono configuration.  This is the most cost effective approach.  As we all understand, when we load the amp with 8ohms (vs 4ohms), the amp voltage output does get a bit higher and that does raise the possibility of over-drive the drivers. 8ohms meaning using only 2 drivers.  I really think 4 drivers or 2 drivers are both close the perfect match in terms of power vs excursion.  We should not over study this pros and cons of 2 drivers (or 3 drivers) vs 4 drivers.   

At the end of last year I did consider EOL to HX800 as the sales of those amps was very slow.  Then an overseas customer want to order more and so I decide to continue HX800. It only comes with one configuration HX800XLR3. 

It takes time to port servo to an arbitrary amp.  I once talked to Icepower people and they don't even want to support me to modify their amps to be servo configured.  I can understand the problem because servo feedback needs to be 100% stable, meaning there is certain phase margin issue to consider.  I am glad I settle with Hypex. It is single-ended and perfect for adding servo control. Plenty of phase margin because the amp already take feedback after LC tank.  I will not entertain another amplifier configuration. The pricing of Hypex is also very reasonable. We sell them at a higher price because the other supporting circuitry and components.  There is no economical incentive to pick another class-D module and start the process again. Hope everyone understands this.

emailtim

Re: DIY conversion of plate amp to servo
« Reply #16 on: 13 Jun 2019, 01:57 am »
SW12-16FR has an DCR of 12ohms.  All of our drivers have taken the approach of 3ohm DCR (4ohm nominal, when one averaging the impedance over frequency).  So if one does the calculation, it takes 4 parallel drivers to make it 12ohm/4 = 3ohms. That was our original goal.  In practice, we do find that A370 (class A/B) series amps, when driving 4 drivers, can lead to amplifier overheat. One can use a fan, or not to play as loud to mitigate the problem.  When driving only 2 SW12-16FR, the A370 series amp is fine.  That amp was introduced more than 10 years ago. Starting 5 years ago, we began to offer HX800 with Hypex class D modules and that completely addresses the amp temperature issue. The next issue is how to put all 4 drivers in close proximity so that servo control remains effective. Since HX800 has  two 400WRMS modules with independent servo control, one can use one module to driver 4 with one module and put it in left and another 4 in right in mono configuration.  This is the most cost effective approach.  As we all understand, when we load the amp with 8ohms (vs 4ohms), the amp voltage output does get a bit higher and that does raise the possibility of over-drive the drivers. 8ohms meaning using only 2 drivers.  I really think 4 drivers or 2 drivers are both close the perfect match in terms of power vs excursion.  We should not over study this pros and cons of 2 drivers (or 3 drivers) vs 4 drivers.   

At the end of last year I did consider EOL to HX800 as the sales of those amps was very slow.  Then an overseas customer want to order more and so I decide to continue HX800. It only comes with one configuration HX800XLR3. 

It takes time to port servo to an arbitrary amp.  I once talked to Icepower people and they don't even want to support me to modify their amps to be servo configured.  I can understand the problem because servo feedback needs to be 100% stable, meaning there is certain phase margin issue to consider.  I am glad I settle with Hypex. It is single-ended and perfect for adding servo control. Plenty of phase margin because the amp already take feedback after LC tank.  I will not entertain another amplifier configuration. The pricing of Hypex is also very reasonable. We sell them at a higher price because the other supporting circuitry and components.  There is no economical incentive to pick another class-D module and start the process again. Hope everyone understands this.

Brian,

Thanks for the information and glad to hear you are still making the HX800XLR3.

rythmik

  • Industry Participant
  • Posts: 181
    • Rythmik Audio
Re: DIY conversion of plate amp to servo
« Reply #17 on: 13 Jun 2019, 12:56 pm »


Back to technical. In my wish to get full excursion from the drivers, I have been calculating on the pretty rough-and-ready basis that the 16R, 200W drivers at least need enough amplifier to get them to 200WRMS, which will require balanced 80V DC rails on the PSU. I would be very, very surprised if the A370 can do this. On the other hand the UcD700 has 90V rails and can do it. The Rythmik HX800 would have been ideal if it was a single module, but if it is 2x400W modules, it can't.

Unless the HX800's two internal modules can be bridged??  :scratch: (dubious look) Even if possible, it would probably be limited to a minimum load of 8R --yes?-- and unable to run 4 drivers.

If someone has a better technical analysis of what it takes in amplification to deliver full excursion from the SW-12-16FR, then I am all ears. Because my calculation above is, how shall I put it kindly, indirect.

cheers
Arg

What if the 200W rating is actually 200W peak.  But the excursion check is best done via a  simulation.   Here is the basspro simulation plot. Note I trick the voltage level by using nominal Z "4ohms", and then enter 400W.  Then it does not matter the actual DCR 12ohms or 14ohms.   The yellow line is when DCR is 14ohms (based on GR spec) and the orange line is for 12ohms (which is what I thought the DCR really is).  Anyway the one way excursion is over 12mm or peak to peak 24mm. 
Since we are on this plot, I can also explain the advantage of servo.  The driver is a high Q by design. As one can see, the excursion at 20hz is almost same as at 10hz (or even 5hz), and at 30hz, it is down by approximately 30-35% only.  If one were to use a low Q value to start with, the 20hz and 30hz achievable excursion when 5hz/10hz excursion is fixed at said 14mm will be much lower.  We do want to achieve a low Q sound and that is done with servo. But we don't want to pay for low Q excursion penalty.  That is why servo OB will be better.   200W rating does not mean anything because these drivers do not reach thermal limit at all. One will bottom out the driver before that happens.  So one lesson for DIY is simulation is very important. Plug in published data and then check simulation results.   
 







tnargs

  • Jr. Member
  • Posts: 51
Re: DIY conversion of plate amp to servo
« Reply #18 on: 13 Jun 2019, 01:51 pm »
Thanks for your comments on the technical side. I am now feeling a lot more encouraged that one HX800 per 4 SW-12-16R drivers actually does deliver enough to get the most out of the drivers.

I must say though, in your simulation above, many parameters are very unlike the SW-12-16R parameters on the GRR web page. Fs, Qms, Vas, Rms, all very different. Is it still a good representation of that driver despite that?

Also, I asked a question about the 8R version of that driver, a few posts above
OK I hadn't even noticed the SW-12-08FR driver. It would certainly fix my amp concerns.

But why is its Fs so much higher?

And what does that mean in practice? (In the ultimate end game, at the highest SPLs and lowest frequencies)

Thanks

Any thoughts on that?

thanks

rythmik

  • Industry Participant
  • Posts: 181
    • Rythmik Audio
Re: DIY conversion of plate amp to servo
« Reply #19 on: 14 Jun 2019, 06:00 am »
Thanks for your comments on the technical side. I am now feeling a lot more encouraged that one HX800 per 4 SW-12-16R drivers actually does deliver enough to get the most out of the drivers.

I must say though, in your simulation above, many parameters are very unlike the SW-12-16R parameters on the GRR web page. Fs, Qms, Vas, Rms, all very different. Is it still a good representation of that driver despite that?

Also, I asked a question about the 8R version of that driver, a few posts above
Any thoughts on that?

thanks

 The T/S parameters difference is very subtle.

8ohm version is for 300WRMS module, not 400WRMS module.  You can still use them with 400WRMS with a bit of over-drive.