Cheap as Chips OPA1688 Low-THD Muscle Amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Founder of XSA-Labs
Joined 2012
Paid Member
SSR Relay DC Protection for HPA

Jhofland designed this handy circuit for protection of DC-coupled headphones using MOSFET solid state relays (SSRs) driven by opto-couplers. Here, a comparator is used to check for DC offsets > +/-500mV after a CRC low-pass filter. I built a verification test circuit and it works very well. Turn on delay is about 2.5 seconds. I tested DC offset shut-down by touching the +/-5v rail from the regulators to the input pin and it shuts off quickly and recovers when the offending DC is removed.

attachment.php


attachment.php


Schematic:
attachment.php
 

Attachments

  • HPA-SSR-Protect-03.jpg
    HPA-SSR-Protect-03.jpg
    392.2 KB · Views: 4,476
  • HPA-SSR-Protect-04.jpg
    HPA-SSR-Protect-04.jpg
    490.5 KB · Views: 4,562
  • HPA-SSR-Protect-Schematic.jpg
    HPA-SSR-Protect-Schematic.jpg
    313.9 KB · Views: 3,923
Founder of XSA-Labs
Joined 2012
Paid Member
Thanks, Abraxalito. That part looks handy - Rdson is a bit higher than the one we are using at present (not sure that matters as much for a headphone). Yes, U5 and U6 are on the underside:
attachment.php


Just took some FFTs with the SSR in and out at 1.00Vpp and 2.00Vpp into 51ohms using the Akitika 2ppm 1kHz source and a Focusrite Solo Gen 2.

1.00Vpp into 51ohms, no SSR (4.2ppm THD):
attachment.php


with SSR (4.4ppm THD):
attachment.php


2.00Vpp into 51ohms, no SSR (9.2ppm THD):
attachment.php


with SSR (9.2ppm THD):
attachment.php
 

Attachments

  • 8x-OPA1622-2.00Vpp-51ohms-Yes-SSR.jpg
    8x-OPA1622-2.00Vpp-51ohms-Yes-SSR.jpg
    257.9 KB · Views: 2,464
  • 8x-OPA1622-2.00Vpp-51ohms-No-SSR.jpg
    8x-OPA1622-2.00Vpp-51ohms-No-SSR.jpg
    257.5 KB · Views: 2,480
  • 8x-OPA1622-1000mVpp-51ohms-Yes-SSR.jpg
    8x-OPA1622-1000mVpp-51ohms-Yes-SSR.jpg
    256.6 KB · Views: 2,555
  • 8x-OPA1622-1000mVpp-51ohms-No-SSR.jpg
    8x-OPA1622-1000mVpp-51ohms-No-SSR.jpg
    258 KB · Views: 2,521
  • HPA-SSR-Protect-05.jpg
    HPA-SSR-Protect-05.jpg
    284.7 KB · Views: 2,511
Last edited:
Thanks, Abraxalito. That part looks handy - Rdson is a bit higher than the one we are using at present (not sure that matters as much for a headphone).

Yes, I noticed that too - there are one or two parts in the family with slightly lower Rdson but they're more expensive and Mouser doesn't have them in stock. The difference is 20mohm for the two Vishay FETs in series vs 30mohm for this Toshiba part (typicals). To put the numbers in context, a 30mm length of 0.5mm wide copper trace has about 30mohm resistance.

Incidentally if you were to revise the board you could save on space (and probably on cost too) if you changed the 1uF film caps to X7Rs. They seem not in a critical position, just doing low-pass filtering prior to the comparators.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Testing out the 8x OPA1622 amp with an eBay dual TPS7A low noise regulator board set for +/-18.5v and an eBay +/-24v DC-DC step up. All powered from 12v Pb-acid gel cell. Source is 2ppm 1kHz Akitika oscillator, and sound card is Focusrite 2i4. Signal out of amp is adjusted to be 1.0Vpp into 51ohm dummy load. Setup is not the cleanest, so picking up some very low level -120dB 60Hz. But overall, the amp with this PSU shows a -130dB noise floor and 5.7ppm THD with dominatn H2 and about -12dB less H3 and nothing else.

attachment.php


attachment.php
 

Attachments

  • Muscle-Amp-w-TPS7A-LDO-24v-DC-DC-1.0Vpp-51ohms-44k.jpg
    Muscle-Amp-w-TPS7A-LDO-24v-DC-DC-1.0Vpp-51ohms-44k.jpg
    237.4 KB · Views: 3,364
  • Muscle-Amp-w-TPS7A-LDO-24v-DC-DC-1.0Vpp-51ohms-44k.-test-setup.jpg
    Muscle-Amp-w-TPS7A-LDO-24v-DC-DC-1.0Vpp-51ohms-44k.-test-setup.jpg
    473.6 KB · Views: 3,284
In a composite amplifier, you need to simulate all loops. Both any inner loops and the outer/global loop. All loops need to be stable. Finding a good point to break the loop for the stability analysis can be tricky, especially if you have multiple loops.

Tom

Tom, you are of course correct on the loop simulation but I don't have enough mortal time remaining to invest on tweaking the cad models.

So I use the 'trial and error' method combined with hunches, guesses and some nice test equipment.

Bruce

-
 
Founder of XSA-Labs
Joined 2012
Paid Member
I’m working with Jhofland on a new Muscle Amp Lite with only 4x OPA1622 for outputs but a much improved front end topology using OPA1656’s. Will have provision for balanced input sources and even BT option. Power will be via a custom DC-DC converter with linear regulators. Looking forward to comparing it with the correct 8x rig.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Here is a teaser shot of the new OPA1656/quad OPA1622 (each package has two opamps). There is a custom balanced input converter, an active Baxandall volume controller, and a SSR DC protection and delay circuit. Also built in regulated power supply. All designed to fit a standard Hammond box.
 

Attachments

  • E1B36741-B815-48D2-92DB-197F056540E9.jpg
    E1B36741-B815-48D2-92DB-197F056540E9.jpg
    150.9 KB · Views: 899
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.