Hi,
Using one Lundahls per psu rail (ie four in total, windings parallel), my sims give something like 18V dc +/- for choke input.This is indeed good enough for the output stage. It might not be enough for the kubota's however. That's where Eduard's suggestion of a separate supply for the input comes in. Given the documentation anything above 25 Vdc will work for kubota's i'd guess.
For now we are still with the rclcrc. So no need to worry about that now. Speaker protection could be considered. I hate those because they tend to put a relay switch in the ls connection. jmho. It might be a good idea though. Never had an amp fail on me so no experience there. You already have a soft start using the cl60's (cl90 gives a softer startup and is an option given230 ac versus 110).
Joris
edit: typo
Using one Lundahls per psu rail (ie four in total, windings parallel), my sims give something like 18V dc +/- for choke input.This is indeed good enough for the output stage. It might not be enough for the kubota's however. That's where Eduard's suggestion of a separate supply for the input comes in. Given the documentation anything above 25 Vdc will work for kubota's i'd guess.
For now we are still with the rclcrc. So no need to worry about that now. Speaker protection could be considered. I hate those because they tend to put a relay switch in the ls connection. jmho. It might be a good idea though. Never had an amp fail on me so no experience there. You already have a soft start using the cl60's (cl90 gives a softer startup and is an option given230 ac versus 110).
Joris
edit: typo
Last edited:
Greetings, and my apologies if I am taking this thread off topic, but I hope it is consistent with the thread.
I am also building a Hiraga Super 30 using Jim's Audio boards. Based on the early discussion and a couple other threads, I would like to try a switching mode psu. My simple minded question is how to connect. The schematic shows a dual rail, +/-24VDC supply but the SMPS is single rail. Do I simply connect the V+ side of the SMPS to the V+ on the amp and the V- side of the SMPS to the V- side of the amp? If so, what about ground?
Thanks for assistance and apologies for showing my ignorance.
I am also building a Hiraga Super 30 using Jim's Audio boards. Based on the early discussion and a couple other threads, I would like to try a switching mode psu. My simple minded question is how to connect. The schematic shows a dual rail, +/-24VDC supply but the SMPS is single rail. Do I simply connect the V+ side of the SMPS to the V+ on the amp and the V- side of the SMPS to the V- side of the amp? If so, what about ground?
Thanks for assistance and apologies for showing my ignorance.
Hi hirscwi - i remember another diy member using smps successfully although if i remember correctly he compared his amplifier to another Hiraga with a traditional PSU (like i am trying to build) and apparently the larger PSU blew the SMPS away.
Ill see if i can find the thread and ill PM you, good luck though let us know how you get on.
You want to base this on the smaller noise threshhold desired for the input stage. Reducing the voltage is simple, re my post #97.
hahaha so basically you are telling me i need bigger?
Large High Quality Heatsink TO220 TO247 and more Transistor Mosfet Heatsink | eBay
how about this?
L : 300m
W: 113mm
H : 82mm
The unit weights about 2.7Kg
how about this?
L : 300m
W: 113mm
H : 82mm
The unit weights about 2.7Kg
Hello,
Buying big power resistors that need big heatsinks or a big slab of aluminium while you dont have these is not very usefull.
I have loads of heatsinks but still i prefer the ceramic power resistors that depend upon air cooling. I bought a few in surplus stores decades ago and recently i dismanteled a few from welding machines at work that needed to be replaced. I have done test with these during 24 hours to be sure everything in the power supply will survive.
3 Ohm 5% 100W Watt High Power Wirewound Resistor U5A6 4894462078884 | eBay
You could get some similar like these. You will have to know the minium dcr value ( the one that will give the highest current draw) No use buying loads of 3 ohm if the miniumum has to be 14 ohm.
By the way i am not sure the diodes on your board will survive. Better have a look at the datasheet to see how much cooling they need.
Bye bye,Ed
Buying big power resistors that need big heatsinks or a big slab of aluminium while you dont have these is not very usefull.
I have loads of heatsinks but still i prefer the ceramic power resistors that depend upon air cooling. I bought a few in surplus stores decades ago and recently i dismanteled a few from welding machines at work that needed to be replaced. I have done test with these during 24 hours to be sure everything in the power supply will survive.
3 Ohm 5% 100W Watt High Power Wirewound Resistor U5A6 4894462078884 | eBay
You could get some similar like these. You will have to know the minium dcr value ( the one that will give the highest current draw) No use buying loads of 3 ohm if the miniumum has to be 14 ohm.
By the way i am not sure the diodes on your board will survive. Better have a look at the datasheet to see how much cooling they need.
Bye bye,Ed
Hi Vishalk,
Yes the resistors for the test load will need a heatsink. I don't see the need to buy an extra heatsink. You have that nice case with large heatsinks. Make a mock up of the psu inside the case and clamp those resistors to the heatsinks. that will do fine.
Eduard has a point about the diodes but that is easily solved by mounting them on the heatsinks of the case. that's the way i did it with the alep-x.
Making a mock up / test setup inside the case gives you some idea how everything is going to fit as well. this helps.
Joris
Edit: had a look at the datasheet for the rectifiers. All is well within the SOA region imho.
Yes the resistors for the test load will need a heatsink. I don't see the need to buy an extra heatsink. You have that nice case with large heatsinks. Make a mock up of the psu inside the case and clamp those resistors to the heatsinks. that will do fine.
Eduard has a point about the diodes but that is easily solved by mounting them on the heatsinks of the case. that's the way i did it with the alep-x.
Making a mock up / test setup inside the case gives you some idea how everything is going to fit as well. this helps.
Joris
Edit: had a look at the datasheet for the rectifiers. All is well within the SOA region imho.
Last edited:
Hi,
for curiosity i've looked at some datasheets comparable to the type of resistors you bought for the test load of the psu. those where the arcol/ohmite ones on mouser.
Dissipation rating without a heatsinks was something like 30W. the 10R resistors will do about 22W and the 8R 18W. So even without heatsinking you should be safe. None the less, clamp them to the heatsinks.
Far safer and reliable that way in terms of temperature and ratings. I gave the advice not to use resistors above 50% of their rating for that reason. clamping them to the heatsinks ensures that advice.
Joris
for curiosity i've looked at some datasheets comparable to the type of resistors you bought for the test load of the psu. those where the arcol/ohmite ones on mouser.
Dissipation rating without a heatsinks was something like 30W. the 10R resistors will do about 22W and the 8R 18W. So even without heatsinking you should be safe. None the less, clamp them to the heatsinks.
Far safer and reliable that way in terms of temperature and ratings. I gave the advice not to use resistors above 50% of their rating for that reason. clamping them to the heatsinks ensures that advice.
Joris
The heathsinks will already get hot with 2 power transistors mounted close together.
The connections between transformer output, diodes and first caps should be short. Using a circuit for the diodes but then mounting the diodes on the case does not sound very logical.
If you use choke input your diodes will have to deliver a kind of constant current with a capacitor input there will charging pulses which will heat up things more!
That is the the reason why with choke input the transformer can deliver more current.
You can do a test with 2 10 ohm resistors in series and tell us how much voltage you will get with a current that will be around 1,1 A . If you will use ll2733 for choke input. Just connect one transformer, two bridge rectifiers, ll2733, 2 caps, 2 10 ohm in series. Stop buying stuff and tell us.
Ed
I hope you do not mind me butting in here, I responded on the other thread, but I am a little confused.
First - I guess you intend to put 18 ohms across each 36 volt rail, if so it will draw 2 amps.
Ohms law tells me that is 72 Watts [36v x 2A = 72 Watts or 40 Watts (10R) and 32 Watts (8R)] per rail. So the heat sink will need to dissipate 144 Watts... I would not do it.
Second - If that is to work out the resistance value of the resistors and choke to get you to 28 volts per rail (from 36 volts) you have to use the 'real' current your amplifier will use. I know nothing about the Hiraga, but I see 1.5 amps quoted, and speculate that is where the 18 ohms is from originally. I.e. 18.6R across 28 volts = a draw of 1.5 Amps. (42 Watts each rail so 84W total.)
So if the 1.5 amps is correct you have to apply that load to your 36 volt supply. Now that requires a resistance of 24 ohms. (Ohms law 36 divided by 1.5 = 24).
Therefore to get 36 volts down to 28 volts at 1.5 amps you need resistance of about 5.4 ohms in the filters (That is 24 - 18.6 = 5.4) That is 12 watts dissipation.
In reality, nothing is quite as straight forward as that. So 5 or 6 ohms is a good place to start.
Sorry if I missed the point, or it has been covered already. Just trying to help.
Alan
First - I guess you intend to put 18 ohms across each 36 volt rail, if so it will draw 2 amps.
Ohms law tells me that is 72 Watts [36v x 2A = 72 Watts or 40 Watts (10R) and 32 Watts (8R)] per rail. So the heat sink will need to dissipate 144 Watts... I would not do it.
Second - If that is to work out the resistance value of the resistors and choke to get you to 28 volts per rail (from 36 volts) you have to use the 'real' current your amplifier will use. I know nothing about the Hiraga, but I see 1.5 amps quoted, and speculate that is where the 18 ohms is from originally. I.e. 18.6R across 28 volts = a draw of 1.5 Amps. (42 Watts each rail so 84W total.)
So if the 1.5 amps is correct you have to apply that load to your 36 volt supply. Now that requires a resistance of 24 ohms. (Ohms law 36 divided by 1.5 = 24).
Therefore to get 36 volts down to 28 volts at 1.5 amps you need resistance of about 5.4 ohms in the filters (That is 24 - 18.6 = 5.4) That is 12 watts dissipation.
In reality, nothing is quite as straight forward as that. So 5 or 6 ohms is a good place to start.
Sorry if I missed the point, or it has been covered already. Just trying to help.
Alan
But we are not connecting the actual amp yet so there will be no transistors to dissipate energy... The 10+8R resistors will simulate the circuit.
Yeah,
I know!! Once i build a 30 watt Hiraga because i had loads of parts i wanted to sell and even with a big heatsink the power transistors will get hot if you mount them close together.
Probably the reason why the original Hiraga had seperate heatsinks for each power transistor.
So no more beating about the bush and wait for the original poster to do some power supply test. Preferably choke input!!! lol
If the actual amp will take as much posts we will all follow Pieter!
Bye bye,Ed
I know!! Once i build a 30 watt Hiraga because i had loads of parts i wanted to sell and even with a big heatsink the power transistors will get hot if you mount them close together.
Probably the reason why the original Hiraga had seperate heatsinks for each power transistor.
So no more beating about the bush and wait for the original poster to do some power supply test. Preferably choke input!!! lol
If the actual amp will take as much posts we will all follow Pieter!
Bye bye,Ed