Here is test with a low THD 8x OPA1622 headphone amp driving the intrinsic impedance in the input stage of the Focusrite Solo (2nd gen) mic input (3kohm). I have a typo on the plot label as 5kohm, but it is 3kohm according to the spec's.
Using this HPA:
Using the Akitika 2ppm 1kHz source, 1.0Vpp into 3kohm load, direct shot no SRR protection circuit in place:
Here is with the SRR protection circuit in betweent the amp and the input stage of the Focusrite:
The THD was 0.00040% for both cases, and there was a little more added higher harmonics (8th) at very low levels (0.000048%). FFT settings were 32k points, 0.94 exp averaging, Blackman-Harris 7 window.
Using this HPA:
Using the Akitika 2ppm 1kHz source, 1.0Vpp into 3kohm load, direct shot no SRR protection circuit in place:
Here is with the SRR protection circuit in betweent the amp and the input stage of the Focusrite:
The THD was 0.00040% for both cases, and there was a little more added higher harmonics (8th) at very low levels (0.000048%). FFT settings were 32k points, 0.94 exp averaging, Blackman-Harris 7 window.
Attachments
Last edited:
A member contacted me via PM asking me to test an extreme case of a low impedance load (say, 3 to 4ohms) and an amp stuck at max rail voltage (say, 48v). It was suggested that this condition might cause the MOSFET or other parts on this protection circuit to overheat and fail.
The test was simple enough to do: I used an 800w 48v SMPS and made a hand momentary switch with two bare wires that I would engage manually. This connected to the input of the SSR protection circuit, and the output goes to a massive 200w x 4ohm dummy load resistor. A DVM is connected to the PSU to make sure it delivers the goods and doesn’t switch off in self-protect, and an O-Scope connected to the dummy load to watch the transient. The protection circuit is powered by 4x9v batteries throughout to make sure it has the continuous power to do the job.
I then manually touched the contacts to close the circuit and each time, the circuit would work and turned off the conduction path and LED would go out as long as the DC held. Reliably, each time the DC was removed, after 3 seconds, the circuit would work again. I repeated this test over 20 times. There was not any sign of heating, funny smells, smoke etc. The circuit continued to work flawlessly. I left the 48vdc connected continuously for 5minutes to see if the board would get too hot. An IR thermometer showed 36C and touching by fingers could barely feel warmth. This was the dissipation in the DC sense resistor which was very well heatsinked to the PCB. The o-scope showed that the circuit reacts and cuts out the DC in 25 milliseconds, thereby preventing any further DC from damaging the speaker cone.
So the short story is, this protection circuit will continue to work and protect your speaker in the event of a full-rail stuck-voltage amp failure. It gets the job done reliably every time.
The test was simple enough to do: I used an 800w 48v SMPS and made a hand momentary switch with two bare wires that I would engage manually. This connected to the input of the SSR protection circuit, and the output goes to a massive 200w x 4ohm dummy load resistor. A DVM is connected to the PSU to make sure it delivers the goods and doesn’t switch off in self-protect, and an O-Scope connected to the dummy load to watch the transient. The protection circuit is powered by 4x9v batteries throughout to make sure it has the continuous power to do the job.
I then manually touched the contacts to close the circuit and each time, the circuit would work and turned off the conduction path and LED would go out as long as the DC held. Reliably, each time the DC was removed, after 3 seconds, the circuit would work again. I repeated this test over 20 times. There was not any sign of heating, funny smells, smoke etc. The circuit continued to work flawlessly. I left the 48vdc connected continuously for 5minutes to see if the board would get too hot. An IR thermometer showed 36C and touching by fingers could barely feel warmth. This was the dissipation in the DC sense resistor which was very well heatsinked to the PCB. The o-scope showed that the circuit reacts and cuts out the DC in 25 milliseconds, thereby preventing any further DC from damaging the speaker cone.
So the short story is, this protection circuit will continue to work and protect your speaker in the event of a full-rail stuck-voltage amp failure. It gets the job done reliably every time.
Attachments
Last edited:
xrk971 - 4 units - USA
pinnocchio - 6 units - CAN
Vunce - 4 units - USA
Miggie78 - 2 units - FINLAND
Nautibuoy - 2 units (maybe 4, depending on price) - UK
gary s - 4 units - AUSTRALIA
WTL- 4 units - AUSTRALIA
ppap64 - 2 units - CANADA
drpro - 4 units - USA
Alarudin - 2 units - SINGAPORE
Ripcord - 2 units - USA
Myint67. - 4 units - Singapore (depends on price )
av-trouvaille - 2 units - NL
ravid - 2 units - India
darrr - 2 units - Poland ( depends on price )
Avtech23 - 2 (maybe 4) units - Australia
bkdog - 4 units - Australia
necplusultra - 2 units - Dominion of Canada
gannaji - 4 units - USA
Amplitude - 2 units - Denmark
esprit - 2 units - UK
hirscwi - 4 units - USA
yoaudio - 4 units - USA
kamco - 6 units - USA
Baobei - 2 units - UK (depends on price)
Kokanee - 4 units - Canada
Abanico - 2 units - Spain
vasilis - 4 units - Greece
astromo - 6 units - Australia
mlloyd1 - 2 units - USA
tony.garner - 2 units - UK (depends on price)
markus22ch - 6 units - Switzerland
redjr - 2 units - USA
gtose - 2 units - USA
jwjarch - 4 units - USA
gktaudio - 6 units - USA
quadtech - 2 units - USA
livingstone - 4 units - RUS (depends on a price)
Bare - 2 units - Can
Quan - 4 units - Aus
MD_Stryker - 2 units - USA
pinnocchio - 6 units - CAN
Vunce - 4 units - USA
Miggie78 - 2 units - FINLAND
Nautibuoy - 2 units (maybe 4, depending on price) - UK
gary s - 4 units - AUSTRALIA
WTL- 4 units - AUSTRALIA
ppap64 - 2 units - CANADA
drpro - 4 units - USA
Alarudin - 2 units - SINGAPORE
Ripcord - 2 units - USA
Myint67. - 4 units - Singapore (depends on price )
av-trouvaille - 2 units - NL
ravid - 2 units - India
darrr - 2 units - Poland ( depends on price )
Avtech23 - 2 (maybe 4) units - Australia
bkdog - 4 units - Australia
necplusultra - 2 units - Dominion of Canada
gannaji - 4 units - USA
Amplitude - 2 units - Denmark
esprit - 2 units - UK
hirscwi - 4 units - USA
yoaudio - 4 units - USA
kamco - 6 units - USA
Baobei - 2 units - UK (depends on price)
Kokanee - 4 units - Canada
Abanico - 2 units - Spain
vasilis - 4 units - Greece
astromo - 6 units - Australia
mlloyd1 - 2 units - USA
tony.garner - 2 units - UK (depends on price)
markus22ch - 6 units - Switzerland
redjr - 2 units - USA
gtose - 2 units - USA
jwjarch - 4 units - USA
gktaudio - 6 units - USA
quadtech - 2 units - USA
livingstone - 4 units - RUS (depends on a price)
Bare - 2 units - Can
Quan - 4 units - Aus
MD_Stryker - 2 units - USA
Last edited:
Hi Folks,
We finally have the pricing info for interested customers. This is for a professionallly-assembled and tested unit. You will have to provide your own connectors (Faston, Molex, etc) or just use the built in through-hole solder pads for direct soldered flying leads. Just connect to a >+24vdc power source from the amp positive rail, connect your amp and your speakers for it to work. The RTR SSR DC protection has been tested with 100w rms continuous load into 8ohms, and tested for reliable protection from a continuous rail DC of 48v. Turn on time delay is 3 seconds.
Price is $40 each - We are now taking pre-orders on my Etsy shop here:
RTR Solid State Relay SRR DC Protection | Etsy
It will still take some time for us to get the production units out - I am estimating 3-4 weeks as we still need to verify first unit of production before authorizing the full run of 120 units. So time to delivery is still TBD. Initial production run will be for 120 units, so please get your pre-order in to reserve yours if still interested.
Thanks for your interest!
Last edited:
I double checked the shipping costs and the cost of shipping is about the same as for previous GBs:
$5 for US plus $1 for each additional unit.
$10 for CAN plus $3 for each additional unit.
$15 for everywhere else plus $3 for each additional unit.
EU cost for stereo pair is $80 + $18 or $98USD.
$5 for US plus $1 for each additional unit.
$10 for CAN plus $3 for each additional unit.
$15 for everywhere else plus $3 for each additional unit.
EU cost for stereo pair is $80 + $18 or $98USD.
xrk971, please update Post #1 and specify the board's physical dimensions. All I've got is a nickel scale but something more precise would be appreciated. I know this detail is somewhere in the thread but this data should be up front anyway. Thanks.
I take it that committing to purchase is best done via etsy?
I take it that committing to purchase is best done via etsy?
Sorry I must have forgotten to mention the PCB size earlier. It is 70mm x 35mm and the M3 screw mounting holes are 62mm x ~26.5mm.
Yes, pre-orders are made by placing an order directly in my shop:
RTR Solid State Relay SRR DC Protection | Etsy
Here is the current interest list:
Yes, pre-orders are made by placing an order directly in my shop:
RTR Solid State Relay SRR DC Protection | Etsy
Here is the current interest list:
Last edited:
Oh. So this circuit is for headphone amps? I thought it was a protection circuit for a speaker amp.
It's (relatively) easy to make a switch that doesn't distort at low current. Making one that also doesn't distort at high current is the trick - and is highly relevant if you're planning to use this circuit as a speaker protection circuit. A high-current test will pick up the distortion of both the MOSFET switches and the PCB layout.
Tom
Hi Tomchr,
It is for speakers, I don't have any ultra-low distortion speaker amps on hand. They are designed for monotonically decreasing harmonic distortion (with frequency) and overall THD will be in the 0.007% to 0.05% levels at high power.
Jhofland did measure the SRR protect circuit with an amp at 100w rms into 8ohms and posted the input signal and output signal on an O-scope and they overlaid perfectly. So perfect that you can't see the yellow trace below the blue trace. This is 80Vpp into 8ohms (or 28.27Vrms), is 10 amps enough current? It's quite apparent that there are no gross or even minor distortions from just visually looking at the perfect match of the traces:
Jhofland doesn't have the FFT measurement setup to do the distorton measurements at present.
If you want, I could send you a sample and you can put it through your nice AP setup and tell us what you think. Let me know.
It is for speakers, I don't have any ultra-low distortion speaker amps on hand. They are designed for monotonically decreasing harmonic distortion (with frequency) and overall THD will be in the 0.007% to 0.05% levels at high power.
Jhofland did measure the SRR protect circuit with an amp at 100w rms into 8ohms and posted the input signal and output signal on an O-scope and they overlaid perfectly. So perfect that you can't see the yellow trace below the blue trace. This is 80Vpp into 8ohms (or 28.27Vrms), is 10 amps enough current? It's quite apparent that there are no gross or even minor distortions from just visually looking at the perfect match of the traces:
Jhofland doesn't have the FFT measurement setup to do the distorton measurements at present.
If you want, I could send you a sample and you can put it through your nice AP setup and tell us what you think. Let me know.
Last edited:
I have plenty of Modulus-286 modules and Modulus-686 boards in stock...
How do you know? From simulation?
That's fantastic! However, even a few percent THD is actually hard to spot on an oscilloscope.
Really? No FFT on the oscilloscope? No sound card (with a resistive divider so the input doesn't fry)?
I'll be happy to do that. Toss me an email with the measurements you need and I'll create a quote for you.
Tom
Indeed, I could use more amps!
Both. Look at Alpha amp thread for example
Aksa Lender P-mos Hybrid Aleph (ALPHA) Amplifier
I thought you were the one who was curious, so offering you a free sample and you can measure to satisfy your curiosity. No quotes or exchange of funds needed.
Last edited:
My approach was more in the spirit of a design review. As in "have you thought of..." or "have you measured..." It appears the answer is: No.
You can build a single Modulus-86 channel for about $120. That would allow you to measure the performance of your circuit (or that of your sound card, depending on which is degrading the measurement the most).
I'll be happy to measure your circuit for you, but I do expect to be fairly compensated for my time. I don't work for free. I don't know anybody who does... My car mechanic charges me $120/hour, for example. Plus shop fees! Audio Precision doesn't calibrate my APx525 for free. Not even if I promise to talk them up on the forums. Such is life. Last I had the AP calibrated, the calibration alone was $1500. The shipping fee was $300!
I think the onus is on the vendor to show that their products work well. You're free to disagree.
Tom