And whatabout burned choke in speaker ? Or falling down wire at box ? 😉 It will be shot, by which you drop " something " to the trunks 😀 .
protection
The trio of output transistors are overrated for the voltage levels they are using, they can stand a short term overload long enough to take out a 'normal' fuse no problem, a 5 amp rail fuse will allow ridiculous output levels and still give a big safety margin...Is it completely bulletproof? probably not, but then neither is an arbitrarily specified, generic V/I limiter... I'm gonna build a dual slope limiter for the higher output model's I'm playing with.
If you factor in the 0.68 ohm emitter resistor acting as an instantaneous current limiter for each transistor, it is pretty clear why no protection on the original amp wasn't a reliability issue...
Stuart
The trio of output transistors are overrated for the voltage levels they are using, they can stand a short term overload long enough to take out a 'normal' fuse no problem, a 5 amp rail fuse will allow ridiculous output levels and still give a big safety margin...Is it completely bulletproof? probably not, but then neither is an arbitrarily specified, generic V/I limiter... I'm gonna build a dual slope limiter for the higher output model's I'm playing with.
If you factor in the 0.68 ohm emitter resistor acting as an instantaneous current limiter for each transistor, it is pretty clear why no protection on the original amp wasn't a reliability issue...
Stuart
Upupa Epops said:
haha! No! I run them at 8v... that keeps them resonable, i can still hear them, but only when it's quiet, i usually have it turned up enough to hide the sound!
Coulomb said:
If that's what you've got, i'd agree with Stuart you should have no problems at all with any of this.... 30vac toroid is "close" to perfect... I'm using 28v, mine is a tightly regulated toroid though...
Aaron
Re: protection
I am going to use rail fuses... Not that i have any at this current moment, but i will.... I've got 5 pairs of output, i wouldn't like to know how much power they could handle!! A lot i'm sure if it was a quick short!
How's your project going Stuart? Haven't heard anything from you in a while!!
Aaron
Stuart Easson said:
I am going to use rail fuses... Not that i have any at this current moment, but i will.... I've got 5 pairs of output, i wouldn't like to know how much power they could handle!! A lot i'm sure if it was a quick short!
How's your project going Stuart? Haven't heard anything from you in a while!!
Aaron
pinkmouse said:
My mistake I meant output protection cricuit, so I should desolder all the 300 series components and leave them blank?
Regards
Anthony
projects...
I've been travelling a bunch, got a new girlfriend, not as local as the last, lots of driving, plus I'm having to go in to the office for work and all manner of other distractions...
But, I made one channel of a ksa50, it worked perfectly, so I started on the other, not quite finished...then I was tempted by k-amps...
He wants a variation on the amp for higher powers. Never one to turn down a challenge I began 'recalibrating' the circuit... I am presently building the output stage, got a nice pair of heatsinks from a deceased adcom gfa565, they hold five TO3 pairs each and I've fitted them together to make a windtunnel. Thermally speaking, fanless, each is just about enough for one channel of a stock ksa50, so I'm gonna use 2 with push-pull 15cm fans for one channel of this amp...I'm quite sure we will need more outputs for 2ohm operation but as I said earlier, I ran out of sockets...
Power supply: Not completely sure yet, but since I have about a bunch of 400va toroids, I think I'm gonna use 4, along with my variac to get unloaded +/- 100v rails, plan B is to use a single 1.4kVA toroid, again with the variac. Currently I have built a couple of the boards with higher voltage small signal transistors throughout, plus moderate heatsinking for the Vas, the increased bias requirements mean the pre-drivers run hot, the drivers are now mj21193/4, though ultimately I think the mjl4281 and it's complement are a better choice.
I'm going to start the idle current low, test everything into a reasonably high impedance, then slowly crank it until the sinks are half way to max. acceptable temp, ~43c...at which point if my theory hasn't gone up in a cloud of magic smoke, we should have an amp with very high output potential... mainly class AB obviously, but still should be able to get the first few tens as class A...I don't have any multi-kilowatt test loads, so I will have to jury rig something, like huge resistor bolted to heatsink immersed in bucket of ice water...
The longest job so far has been bolting and soldering all the driver and output stuff together. I am triple checking everything, a failure in this 'part' will be spectacular...safety goggles anyone...ear defenders..., kevlar vest, and as I said the fire extinguisher is close by...
Stuart
I've been travelling a bunch, got a new girlfriend, not as local as the last, lots of driving, plus I'm having to go in to the office for work and all manner of other distractions...
But, I made one channel of a ksa50, it worked perfectly, so I started on the other, not quite finished...then I was tempted by k-amps...
He wants a variation on the amp for higher powers. Never one to turn down a challenge I began 'recalibrating' the circuit... I am presently building the output stage, got a nice pair of heatsinks from a deceased adcom gfa565, they hold five TO3 pairs each and I've fitted them together to make a windtunnel. Thermally speaking, fanless, each is just about enough for one channel of a stock ksa50, so I'm gonna use 2 with push-pull 15cm fans for one channel of this amp...I'm quite sure we will need more outputs for 2ohm operation but as I said earlier, I ran out of sockets...
Power supply: Not completely sure yet, but since I have about a bunch of 400va toroids, I think I'm gonna use 4, along with my variac to get unloaded +/- 100v rails, plan B is to use a single 1.4kVA toroid, again with the variac. Currently I have built a couple of the boards with higher voltage small signal transistors throughout, plus moderate heatsinking for the Vas, the increased bias requirements mean the pre-drivers run hot, the drivers are now mj21193/4, though ultimately I think the mjl4281 and it's complement are a better choice.
I'm going to start the idle current low, test everything into a reasonably high impedance, then slowly crank it until the sinks are half way to max. acceptable temp, ~43c...at which point if my theory hasn't gone up in a cloud of magic smoke, we should have an amp with very high output potential... mainly class AB obviously, but still should be able to get the first few tens as class A...I don't have any multi-kilowatt test loads, so I will have to jury rig something, like huge resistor bolted to heatsink immersed in bucket of ice water...
The longest job so far has been bolting and soldering all the driver and output stuff together. I am triple checking everything, a failure in this 'part' will be spectacular...safety goggles anyone...ear defenders..., kevlar vest, and as I said the fire extinguisher is close by...
Stuart
Re: projects...
What is the process of adjusting the three 5K pots and what are those two little pads for marked High Low?
Regards
Anthony
Stuart Easson said:
What is the process of adjusting the three 5K pots and what are those two little pads for marked High Low?
Regards
Anthony
Re: protection
The quadruple octet of output trannies I'm using has 2 Ohm for Re (as per KMA 200). I think it is a good idea to use high(ish) values there as it enforces (rather than promotes) current sharing pretty efectively. It also limits the current per transistor pretty much to such levels that either even a large fuse will blow well before transistor (rated @25A) goes.
There's no downside to using high values of emiter resistor that I see except from a miniscule loss to damping factor (who cares!) and a tiny loss of efficiency (which if you worry you shouldn't be building a class A amp in the first place 😀 ).
Bratislav
Stuart Easson said:
The quadruple octet of output trannies I'm using has 2 Ohm for Re (as per KMA 200). I think it is a good idea to use high(ish) values there as it enforces (rather than promotes) current sharing pretty efectively. It also limits the current per transistor pretty much to such levels that either even a large fuse will blow well before transistor (rated @25A) goes.
There's no downside to using high values of emiter resistor that I see except from a miniscule loss to damping factor (who cares!) and a tiny loss of efficiency (which if you worry you shouldn't be building a class A amp in the first place 😀 ).
Bratislav
Re: Re: projects...
You only "need" 2 pots - one of the bias ones is for "high/low" setting- quite literally for hotter climates (like mine!) to have a switch for really hot days to lower the heat level.
Set them about the 'middle' and then start moving them around till you get it right!!! 🙂
Aaron
Coulomb said:
You only "need" 2 pots - one of the bias ones is for "high/low" setting- quite literally for hotter climates (like mine!) to have a switch for really hot days to lower the heat level.
Set them about the 'middle' and then start moving them around till you get it right!!! 🙂
Aaron
You really need two meters. One is set to measure DC offset from output to ground, and the other measures voltage drop across one of the output resistors. Short the input to ground, then adjust the bias pot to give you say 200 mV across the resistor. Then adjust the offset pot to zero. Leave for half an hour or so for things to stabilise, and check that your heatsinking is working properly, then increase the bias 100mV at a time, checking offset each time, and again leaving a period of time for things to bed in before proceeding.
Once you have bias where you want it, run a signal through the amp, (loaded with a dummy load), for a while and then measure again, to make sure things are stable.
Then connect up a test driver, and listen to some music! 🙂
Once you have bias where you want it, run a signal through the amp, (loaded with a dummy load), for a while and then measure again, to make sure things are stable.
Then connect up a test driver, and listen to some music! 🙂
pinkmouse said:
Thanks for being specific, i was at work and in a rush 🙂
2 meter's is invaluble... I actually "started" by hooking up the board only and leaving the outputs themselves disconnected from the rails (so no damage could occur) and powered up the board to make sure there was no shorts, etc... Lucky i did as one of my boards had a solder dag which was shorting!!
Aaron
Alrodent,
can you tell me the spacing between the output devices, from collector pad to the next collector pad ?
I'd like to know if i can use big Sanken's instead of TO247's, 12 on 2 output boards.
can you tell me the spacing between the output devices, from collector pad to the next collector pad ?
I'd like to know if i can use big Sanken's instead of TO247's, 12 on 2 output boards.
Thanks Al,
should be enough, the MT200's are 36mm, if i bend the legs of the outer devices 1.5mm off center they'll be in line on the boards.
(aquired a number of 0.27 C/W heatsinks, with a couple of Papst vents and 12 devices per channel, 200 watts class A in 2 Ohms, on 25vac toroids, 47.5 C heatsink temperature at 25 ambient )
should be enough, the MT200's are 36mm, if i bend the legs of the outer devices 1.5mm off center they'll be in line on the boards.
(aquired a number of 0.27 C/W heatsinks, with a couple of Papst vents and 12 devices per channel, 200 watts class A in 2 Ohms, on 25vac toroids, 47.5 C heatsink temperature at 25 ambient )
Jacco, are in Netherland energy for free ? 200 W / A class is cca 1200 W of continue power consumtion, or are you rich man ? 😉
As a side note, i'm noticing in my room with the doors open (to the rest of the house) i'm getting about 3deg rise in room temp!!! It's not a "small" room, it's 7x5.5m.... Big enough..... With the doors closed it's a nice 22deg 🙂 Outside (next room) is 19deg 🙂 Big difference in temp..... I like these amps now... i might get working on the bias switch for summer though!
Aaron
Aaron
