Hey guys,
I want to switch to an active crossover ciruit and need your help choosing the right amp board. The crossover frequency will be around 1900Hz.
Current setup
Do you think it makes sense to use two extra Honeybadgers for the Tweeters or should i get another amp for them?
It would be nice to get a similar specced amp (apart from output power) that is smaller in size and/or cheaper to build.
Any recommendations are helpful and appreciated!
Cheers,
Martin
I want to switch to an active crossover ciruit and need your help choosing the right amp board. The crossover frequency will be around 1900Hz.
Current setup
- 2x Dayton Audio RST28F-4
- 2x Dayton Audio RS180S-8
- 2x DiyAB Honeybadger
- DIY DAC and preamp
Do you think it makes sense to use two extra Honeybadgers for the Tweeters or should i get another amp for them?
It would be nice to get a similar specced amp (apart from output power) that is smaller in size and/or cheaper to build.
Any recommendations are helpful and appreciated!
Cheers,
Martin
One of the things that i learned from Julian Vereker still sticks with me is “use the best amp on top”. In his specific example if one has active Lin Isobariks and you have 2 NAIM 160 and a 250, you use the 250 up top.
A small class A amplifier would be suggestted. An inexpensive one would be the ACA.
But you might also want to get a better tweeter. I am pretty sure i have a set of those Daytons in my clearance pile ($20/pr).
dave
A small class A amplifier would be suggestted. An inexpensive one would be the ACA.
But you might also want to get a better tweeter. I am pretty sure i have a set of those Daytons in my clearance pile ($20/pr).
dave
I have leftover PCBs for a "Baby"-Aksa:
Based on Hugh Dean's AKSA 55
Partially SMD though.
Also made a design for this tiny class-A but unfortunately there is no leftovers:
Class A, 4 Watt, No Feedback, Simple Circuit, Great Sound
Have a look at the "swap meet" part of the forum.
There is lots of good deals from people selling their leftovers.
Maybe you should state what kind of power supply you intend using or is that going to be a separate chassis?
Based on Hugh Dean's AKSA 55
Partially SMD though.
Also made a design for this tiny class-A but unfortunately there is no leftovers:
Class A, 4 Watt, No Feedback, Simple Circuit, Great Sound
Have a look at the "swap meet" part of the forum.
There is lots of good deals from people selling their leftovers.
Maybe you should state what kind of power supply you intend using or is that going to be a separate chassis?
I´m sure that is a sweet amp, why not? Maybe ask if you can lower the rail voltages with no ill effects.
Last edited:
Thanks for the suggestions!
Totally forgot to mention, I have a 600VA 2x 40V-0-40V Transformer. The plan in to integrate the new amps in my existing chassis and power them from this transformer
Totally forgot to mention, I have a 600VA 2x 40V-0-40V Transformer. The plan in to integrate the new amps in my existing chassis and power them from this transformer
For a high-frequency speaker, you can make an amplifier with a unipolar supply of + 56V and a capacitor at the output.
Sure, but i already have a bipolar supply.
To clarify, I use 4 rectifier and cap multiplier ciruits and have 2x +/- 50V DC supply rails.
With the same transformer, two diodes and an electrolytic capacitor, you get an additional power supply 0 (ground) ... + 56v 🙂
Choose a high yield tweeter like FT66H FOSTEX https://www.fostexinternational.com/docs/speaker_components/pdf/ft66hrev.pdf and drive it with RC4556 built in unity gain follower. Unity gain amplifier or voltage follower in a voltage divider
RC4556 Datasheet PDF ( Pinout ) - High Output Current Dual Operational Amplifier
a FT66H tweeter needs only 100 mV at full sound level and RC4556 can drive 8 Ohm at 100 mV level.
I = U/R = 100mV/8Ohm = 12 mA and the limit is 70 mA.
RC4556 Datasheet PDF ( Pinout ) - High Output Current Dual Operational Amplifier
a FT66H tweeter needs only 100 mV at full sound level and RC4556 can drive 8 Ohm at 100 mV level.
I = U/R = 100mV/8Ohm = 12 mA and the limit is 70 mA.
Attachments
OP, are you using a waveguide for the tweeter? 7" + 28 mm can be made to work but tends to be a bit borderline otherwise, and it would help bring up efficiency near crossover so the targeted 1900 Hz XO becomes more viable.
I hope you are well aware of the challenges to grounding brought about by running multiple amplifiers off a single transformer.
I can see two possible approaches:
1. Use the available +/-50 V supplies to power your tweeter power amps and run tweeter with a series dropper resistor. If you manage to keep sensitivity to 95 dB across the band, you can afford at least a drop of 8 dB (the woofer is at ~87 dB), so the amplifier circuitry used would have to be able to drive maybe 8 ohm loads minimum. Given the limited amount of spectral energy above 1.9 kHz, you are likely to get away with even more.
2. Use a small additional transformer to feed a smaller, low-noise power amp of lower gain, basically a beefed-up headphone amplifier that can drive a 4 ohm tweeter directly. +/-20 V (unreg) and +/-15 V (reg) could be used to power an NJM4556A driving a pair of power BJTs (maybe D44H11/D45H11?) with a transistor bias spreader (maybe use a BD139 for that), plus the usual opamp feedback cap, output inductor and Zobel. This opamp was designed to drive 150 ohm headphone loads, so it has more oomph than almost everything else. Slew rate is a tad on the low side, but remember that lower output voltage means correspondingly reduced demands in this regard as well. (If all else fails, going to power Darlingtons or MOSFETs would enable use of many common opamps.)
I hope you are well aware of the challenges to grounding brought about by running multiple amplifiers off a single transformer.
I can see two possible approaches:
1. Use the available +/-50 V supplies to power your tweeter power amps and run tweeter with a series dropper resistor. If you manage to keep sensitivity to 95 dB across the band, you can afford at least a drop of 8 dB (the woofer is at ~87 dB), so the amplifier circuitry used would have to be able to drive maybe 8 ohm loads minimum. Given the limited amount of spectral energy above 1.9 kHz, you are likely to get away with even more.
2. Use a small additional transformer to feed a smaller, low-noise power amp of lower gain, basically a beefed-up headphone amplifier that can drive a 4 ohm tweeter directly. +/-20 V (unreg) and +/-15 V (reg) could be used to power an NJM4556A driving a pair of power BJTs (maybe D44H11/D45H11?) with a transistor bias spreader (maybe use a BD139 for that), plus the usual opamp feedback cap, output inductor and Zobel. This opamp was designed to drive 150 ohm headphone loads, so it has more oomph than almost everything else. Slew rate is a tad on the low side, but remember that lower output voltage means correspondingly reduced demands in this regard as well. (If all else fails, going to power Darlingtons or MOSFETs would enable use of many common opamps.)
How about a pair of simple buck converters to drop it to +/-20 volts? You can probably find single chip IC solutions (just add LC filter). Run the “super headphone amp” off of that. An op amp driving a couple of CFPs with D44/D45 outputs does sound pretty damned good. BTDT.
Grounding everything wouldn’t be any harder than any other 4 channel amplifier. It would certainly be more of an issue trying to run the tweeter amp off just the + supply. That would combine signal and power grounds, which is what leads to trouble.
Grounding everything wouldn’t be any harder than any other 4 channel amplifier. It would certainly be more of an issue trying to run the tweeter amp off just the + supply. That would combine signal and power grounds, which is what leads to trouble.
I'm not using a waveguide at the moment.
Dayton quotes the efficiency as db @ 2.83V/1m, so 86.9dB @1W/1m for the Woofer and 93.5dB @2W/1m = 90.5dB @1W/1m for the Tweeter. Means if i want do drive the Woofer at ~60W, the Tweeter needs about 26W.
I have +/- 15V (reg) or a little bit more unregulated from my preamp with plenty of power headroom. I could use that voltage to drive an OPA1622 and a pair of BJTs. But that wouldn't give me full output power, I'd need about 10V RMS for 26W @ 4 Ohm.
Maybe to consider for the future
Just use the +/15, and run the power transistors off the unregulated supply. It will be enough power. You’ll get 9 volts rms instead of 10. Whoopee. Wasting a handful of volts in the outputs isn’t the end of the world thermally on something this small, and it reduces the load on the regulator.
If you do end up using the “op amp driving a power stage “ approach, run the op amp in class A. Select an op amp which tends to go to the positive rail with the inputs grounded (offset phase can be either, so test it). Connect the bias spreader from the output to a current source (about 4mA) to the negative rail. Add CFP or darlington output stage and you’re good to go. CFP will get you the most output swing, and the D44 type transistors have the lowest Vce(sat) of anything practical here. Plenty fast. It will be better than a “gainclone” since the thermal feedback from the output stage to the diff pair has been removed, and you can use a *really* nice op amp. You could get adventurous and use an output stage with a little bit of gain instead of a follower, but then it might need more than just a zobel and a lead compensation cap across the feedback resistor to make stable.
If you do end up using the “op amp driving a power stage “ approach, run the op amp in class A. Select an op amp which tends to go to the positive rail with the inputs grounded (offset phase can be either, so test it). Connect the bias spreader from the output to a current source (about 4mA) to the negative rail. Add CFP or darlington output stage and you’re good to go. CFP will get you the most output swing, and the D44 type transistors have the lowest Vce(sat) of anything practical here. Plenty fast. It will be better than a “gainclone” since the thermal feedback from the output stage to the diff pair has been removed, and you can use a *really* nice op amp. You could get adventurous and use an output stage with a little bit of gain instead of a follower, but then it might need more than just a zobel and a lead compensation cap across the feedback resistor to make stable.
Thanks for the tips, i'll give it a try!
Whats the reason I should look for this op amp behavior?
A smaller turn-on transient. If you use this arrangement (with feedback applied) the op amp output will normally sit at a slightly positive voltage, about +2Vbe (with a 2EF output stage) with no signal. The op amp will find it “easier” to properly center if it’s input offset normally sends it to the positive rail rather than the negative rail when running open loop.
You could point the current source in the other direction, and have the output sit at -2Vbe. Then you would want the opposite phase. But op amps usually have a better NPN output than its PNP, the transistors have a higher fT. So it is preferable to have the better side active. There are other arrangements of coupling to drive the output stage symmetrically, but the are all more complicated and dont work as well as the simple trick to force the op amp into class A. The only thing to be mindful of is power dissipation. Any op amp will drive a couple milliamperes - you might not be able to ask more than 10 or 15 mA. That’s plenty to drive a pair of 2EF or CFP.
You could point the current source in the other direction, and have the output sit at -2Vbe. Then you would want the opposite phase. But op amps usually have a better NPN output than its PNP, the transistors have a higher fT. So it is preferable to have the better side active. There are other arrangements of coupling to drive the output stage symmetrically, but the are all more complicated and dont work as well as the simple trick to force the op amp into class A. The only thing to be mindful of is power dissipation. Any op amp will drive a couple milliamperes - you might not be able to ask more than 10 or 15 mA. That’s plenty to drive a pair of 2EF or CFP.