Phase: I did some tests on 82vdc finally. Ran two channels for about one hour, with a target bias = 50ma or 12.2mv across the ERs, the resistor value you want is 2.575k
Initially with cold sinks, it will read lower. However with about 7 vrms driving into 8 ohms for 40 minutes, and the sinks at about a mild 90F, it should track to 10-12mv
Initially with cold sinks, it will read lower. However with about 7 vrms driving into 8 ohms for 40 minutes, and the sinks at about a mild 90F, it should track to 10-12mv
Have you figured out what duty the two 47uf caps in the center of the board serve?
Was wondering if they were part of the feedback loop somehow, they are rated at 63v.
You posted as I was just writing here; thanks for that info, that should be perfect then at around 2.5k - 2.6k if I can find that value. I will get that this next week sometime.
Thanks again...
How does it sound now?
Was wondering if they were part of the feedback loop somehow, they are rated at 63v.
You posted as I was just writing here; thanks for that info, that should be perfect then at around 2.5k - 2.6k if I can find that value. I will get that this next week sometime.
Thanks again...
How does it sound now?
Last edited:
Sounded better with high bias, but that will be another project to close match the OPs. Currently sounds a little thin, but bass is tight and clean.
I hated the headers it came with to connect the supplies since that requires pigtail cables that were not supplied with the kit. So I installed screw in bus types.
I think you are right on the 47 uf caps.... not sure what values should be used but a 47 uF WIMA will be HUGE! I read somewere that those might affect sound... I can consider replacing with films, but they will be expensive and the size of soda cans. Might consider that as a later project.
I hated the headers it came with to connect the supplies since that requires pigtail cables that were not supplied with the kit. So I installed screw in bus types.
I think you are right on the 47 uf caps.... not sure what values should be used but a 47 uF WIMA will be HUGE! I read somewere that those might affect sound... I can consider replacing with films, but they will be expensive and the size of soda cans. Might consider that as a later project.
I was thinking that it might be worth trying some Nichicon muse bp that I have in that value, to see if there would be a difference. I agree that a film cap would be way too large.
Good move on using a terminal strip for the power leads, I have soldered mine.
I was hoping the top end would clear up with a little bias applied, the bass is very nice indeed.
Maybe I will have to settle for what I get with the 50ma or so...
Good move on using a terminal strip for the power leads, I have soldered mine.
I was hoping the top end would clear up with a little bias applied, the bass is very nice indeed.
Maybe I will have to settle for what I get with the 50ma or so...
I have some ELNA Silmic 100uF's I could try... but given that I am currently planning for these 8 channels (4 kits) to be bridged and run 4 15" subs in my HT set up, I am not too concerned about the highs...
I did a few power tests today and blew a coupe of undersized fuses in the process (cheap insurance). Using the Adcom GFA-555ii Transformers I have, which gives me about 81.5vdc rails: I got a the following readings at clipping into 6 ohms with a 315Hz sinewave test tone:
Single channel clipping @ 6 ohms = 51.8vrms = 447 watts
2 channels clipping @ 6 ohms = 49.3vrms = 405 watts
Tomorrow will try 8 ohms and 4 ohms. The wife had enough of me driving her bonkers... plus blew a couple of 4A rail fuses ... need more to continue testing. Eventually once all is secure, i will install 7A rail fuses and a 15A mains fuse.
I did a few power tests today and blew a coupe of undersized fuses in the process (cheap insurance). Using the Adcom GFA-555ii Transformers I have, which gives me about 81.5vdc rails: I got a the following readings at clipping into 6 ohms with a 315Hz sinewave test tone:
Single channel clipping @ 6 ohms = 51.8vrms = 447 watts
2 channels clipping @ 6 ohms = 49.3vrms = 405 watts
Tomorrow will try 8 ohms and 4 ohms. The wife had enough of me driving her bonkers... plus blew a couple of 4A rail fuses ... need more to continue testing. Eventually once all is secure, i will install 7A rail fuses and a 15A mains fuse.
So, a resistance of 2,530 ohms was tried, and the voltage over the emitters began to rise over a two minute period to 23mv when the plug was pulled and the experiment concluded.
I think that with a rudimentary bias scheme, as supplied on this board, the best I will get is around maybe 40ma at the transistors...
Doesn't sound all that bad either, may try to see what I can gain by replacing that 47uf next.
I think that with a rudimentary bias scheme, as supplied on this board, the best I will get is around maybe 40ma at the transistors...
Doesn't sound all that bad either, may try to see what I can gain by replacing that 47uf next.
This is why I installed pots. You may want to try 2.47k. Every channel will vary based on the device parameters. 23mv on unmatched devices is high unless it settles there. You woyld need a 1-2 hour window to check this.
I am assuming the lone d669 is mounted on the same heatsink. Better yet mount it on one of the OP devices if it does not reign in bias
I am assuming the lone d669 is mounted on the same heatsink. Better yet mount it on one of the OP devices if it does not reign in bias
Yeah, I could mount it to the hottest one!
There is some loctite, two part adhesive that is thermally conductive that would work for that...
I may try some lesser value and see if it behaves. It may theoretically settle in time, however, it's what it does during that (delay) time that isn't looking so good at the moment. I could let it run a bit and see if it heats up/settles down I suppose, hmmm.
The drifting was only on one group of outputs, where with less biasing, there wasn't this drifting, so that is bothersome.
That must be some incredible bass that you have planned with multiple channels of these!
There is some loctite, two part adhesive that is thermally conductive that would work for that...
I may try some lesser value and see if it behaves. It may theoretically settle in time, however, it's what it does during that (delay) time that isn't looking so good at the moment. I could let it run a bit and see if it heats up/settles down I suppose, hmmm.
The drifting was only on one group of outputs, where with less biasing, there wasn't this drifting, so that is bothersome.
That must be some incredible bass that you have planned with multiple channels of these!
Yes, and the secret is to partner a high xmax Sub with a low Xmax accordian type surround woofer like the jbl 2035hpl . 4 in total, 2 subs and 2 jbls. Thats why I needed 4 channels.
The 2035's are a great steal since everyone focuses on the $600 2226 series... you can grab a 2035hpl cheap!!!
Back to the topic, I have 2 10 ohm 250watt power resistors, I need to tap it so that it becomes 8 ohms for power tests... let me work on that and report back my 4 ohm and 8 ohm power tests...
The 2035's are a great steal since everyone focuses on the $600 2226 series... you can grab a 2035hpl cheap!!!
Back to the topic, I have 2 10 ohm 250watt power resistors, I need to tap it so that it becomes 8 ohms for power tests... let me work on that and report back my 4 ohm and 8 ohm power tests...
I ran 2 kits (4 channels on the same massive heatsink) for about 2-3 hours. 2.575k is the sweet spot for bias. After warm up, the readings stay between 9mv-12mv. The d669 tracks them ok, with a little thermal inertia, but does track and bring it down eventually if it goes high
Next step, bridging. I will attempt a 555ii style connection. All I need is a switch, a wire and 22k resistor. I will drive channel B from the output of channel A with the 22k resistor going into the feedback loop of channel B. since the fb is inverting, with a zero input on its input, it will invert all of the signal of channel A and replicate it, giving me an out of phase output on channel B and in phase on channel A. Connecting the 2 hot terminals of channel A and B should give me the bridged output.
Another option is to drive both channels from a single XLR outputting crossover, the tip and ring are out of phase and should be able to be used as inputs on 2 channels...
more later.
Next step, bridging. I will attempt a 555ii style connection. All I need is a switch, a wire and 22k resistor. I will drive channel B from the output of channel A with the 22k resistor going into the feedback loop of channel B. since the fb is inverting, with a zero input on its input, it will invert all of the signal of channel A and replicate it, giving me an out of phase output on channel B and in phase on channel A. Connecting the 2 hot terminals of channel A and B should give me the bridged output.
Another option is to drive both channels from a single XLR outputting crossover, the tip and ring are out of phase and should be able to be used as inputs on 2 channels...
more later.
Last edited:
That's the amount that I was shooting for initially, thanks for the handholding there...
I did let it run at 2475ma and it did manage to correct itself, so thanks again.
The one channel is showing 7ma with that same value, the other at 14ma, so will do some fine tuning here. I had started to get some pots, however there wasn't a good source nearby, so I started playing with resistors instead.
I had noticed the pre-drivers getting warm back when I first powered it up for the first time, but had largely forgot about that by now. Now that I'm feeling better about this amp being a worthy piece, I will make up some sheet metal to attach to those, thanks for the reminder
I did let it run at 2475ma and it did manage to correct itself, so thanks again.
The one channel is showing 7ma with that same value, the other at 14ma, so will do some fine tuning here. I had started to get some pots, however there wasn't a good source nearby, so I started playing with resistors instead.
I had noticed the pre-drivers getting warm back when I first powered it up for the first time, but had largely forgot about that by now. Now that I'm feeling better about this amp being a worthy piece, I will make up some sheet metal to attach to those, thanks for the reminder
You are very welcome.
The predrivers in the 555ii also ran very hot and survived for 20+ years but then Began to blacken the pcb and short out. I fixed many of them. Thats why I did it on this guy.
There will be variation in individual devices on the OP, so it does require a pot to fine tune. The variation is time and temperature dependant as well, another variable. Having said that my 4 channels varied from 2.49k to 2.61k to be stable at 12mV (48mA).
I will now begin drilling the heatsink for the next two boards today....
This is the one I used. http://m.ebay.com/itm/lot-20-trimme...%3A33f381b015e0a99be16dfb29fffec824%7Ciid%3A7
The predrivers in the 555ii also ran very hot and survived for 20+ years but then Began to blacken the pcb and short out. I fixed many of them. Thats why I did it on this guy.
There will be variation in individual devices on the OP, so it does require a pot to fine tune. The variation is time and temperature dependant as well, another variable. Having said that my 4 channels varied from 2.49k to 2.61k to be stable at 12mV (48mA).
I will now begin drilling the heatsink for the next two boards today....
This is the one I used. http://m.ebay.com/itm/lot-20-trimme...%3A33f381b015e0a99be16dfb29fffec824%7Ciid%3A7
Last edited:
After some "adjusting" of the values, am right about where I want to be now.
I am using the crappy sil-pads that came with the board, and I think that they work better now after a heat cycle, hadn't been anything but cold until the other day.
I will check the torque on the fasteners after a few more.
I had worked with arrays of hand assembled leds doing some conceptual work years ago and have been aware/paranoid regarding the thermal/bussing thing whenever that is on the table.
The amp sounds pretty nice now, will be headed to its forever home in a party house here soon after some more testing and the heat sinks on the pre drivers. It's future owner is actually a very discerning type, but you'd never know it from a casual glance, and has been very good to me over the years. My plan is to give him this ordinary looking pa amp case with something better inside. The huge power will be a plus.
I am using the crappy sil-pads that came with the board, and I think that they work better now after a heat cycle, hadn't been anything but cold until the other day.
I will check the torque on the fasteners after a few more.
I had worked with arrays of hand assembled leds doing some conceptual work years ago and have been aware/paranoid regarding the thermal/bussing thing whenever that is on the table.
The amp sounds pretty nice now, will be headed to its forever home in a party house here soon after some more testing and the heat sinks on the pre drivers. It's future owner is actually a very discerning type, but you'd never know it from a casual glance, and has been very good to me over the years. My plan is to give him this ordinary looking pa amp case with something better inside. The huge power will be a plus.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.