Stress testing a car SMPS?

[MikeHunt79]

I've recently butchered an old car amp for it's SPMS, and I was wondering if there was any way of finding out how much current it is capable of?

I was planning on hooking up an amp module, putting a sine thru the amp, and cranking up the volume... but I guess this might result in some magic smoke being released...

EDIT: Piccy:
[IMGDEAD]http://img207.imageshack.us/img207/5145/n699581380370457679hx4.jpg[/IMGDEAD]

Does this spms look like it would be capable of 10a on one of the rails without too much sag?

 

[luka]

Hi

I don't think that battery will give 10A and have 12v on it... If you put sinewave on amp this will be max stress for amp, probably smps too, but I load smps alone only, so peak power with amp is in my case all the time, so if it works like that it won't have any problem with driving amp

Do you have any 12v charger, or some bigger battery? If you have car, you could do this test in car/or just outside of it too, since you have two fuses on amp.... you could do test with one fuse only if it is rated more then 10A...but be sure that where will be going just 10A into it, I don't know that amps you have but sinewave will be brutal, so test it step by step, not full volume on the first go

BTW:Smps is/looks big enough for lot more then just 10A

 

[MikeHunt79]

Nice, that's good to know. I was only using the small SLA batt in the pic above to test my rail voltage. I cannot have it higher than 28v... I wasn't planning on using it for stress testing, it's only a 2AH unit.

As for the fuses, the amp came with 2 x 25amp fuses fitted, as as I'm only using one of the rails, 25amp with 12v is more than it will draw when outputting 10A at 24v... I just want to be sure it doesn't kill my amp module, hence the testing.

When stress testing it, I'll be using my car battery, as this is more up to the job. I've had a slightly better idea than using a sine thru the amp module tho, I have quite a few 12v bulbs.

So if I was to parallel these bulbs in pairs, and keep adding them until I reach 10 amps, would this also work? each bulb is around 25w, so 10 pairs of them will pull just over 10A at 24v.

 

[MikeHunt79]

Ok, I've done some testing using my car battery...

No load - 13.55v in - 23.81v out
ratio: 1:1.757

4.5A load (2 bulbs) - 12.38v in - 20.33v out
ratio: 1:1.642

9.2A load (4 bulbs) - 11.91v in - 18.37v out
ratio: 1:1.542

Now, this is the first SMPS I have tested, but doesn't the rail voltage appear to sag quite a bit? Is this due to the toroid or could the rest of the circuit be causing this?

There are only 2 x 1000uF caps for each rail, so would a cap upgrade help here, or do caps only help during transients?

OR...

Could I use both secondaries in parallel, instead of in series like they are right now, making use of both of the secondaries?

My amp module only needs 24v, not +-24v.

 

[luka]

Hi

Well transformer will sag, but see how much your input voltage goes down, in car you will have 13.8V almost for sure 100% of the time, since this is not too big load. In car, your smps would probably have above 21V, and if you take into acount that you won't be able to load your smps with that brutal load, but only amp, it probably will be 23v+

Upgrading cap won't help, since that 1000uF @ say 30kHz is like you would have 1F @ 50Hz, you can use 2200uF but you would need to use scope to see how much ripple do you even have

Yea do parallel sec, that will help, since full period will be used and more wire....

 

[MikeHunt79]

Ahh, I forgot about battery voltage sag... Assuming 13.8v constant input and a 9.2A load, I should get 21.3v which is fine.

I don't think I will bother paralleling the secondaries after all, as I need to make an op-amp based crossover which need +-15v rails which I totally forgot about... Also, Nothing really got very hot on the SMPS which is good, even when pulling 9a for around 10 mins... It seems to be up to the task.

I've got a scope I can test for ripple with, but I'm using a class D module which sounds good during the tests I have done, so I don't think I'll bother... This SMPS runs at 27kHz FWIW...

 

[ifrythings]

This SMPS runs at 27kHz

You mite consider sync'ing the smps to your class D amp so you won't have as much of a posibility of beat frequencies.

BTW I would use both secondaries of the transformer, it's a better use of the transformer and will require less current from each secondary (and it's not hard to do).

 

[MikeHunt79]

It powers H4 bulbs real well:
[IMGHTTPDEAD]http://img338.imageshack.us/img338/809/imgp1083fg5.jpg[/IMGHTTPDEAD]

You mite consider sync'ing the smps to your class D amp so you won't have as much of a posibility of beat frequencies.

Well, I must admit, I wouldn't have a clue how to do this... I've measured ripple on the speaker outputs of my class-d module of around 330KHz, so this might be a bit too high frequency for the SMPS to run at...

BTW I would use both secondaries of the transformer, it's a better use of the transformer and will require less current from each secondary (and it's not hard to do).

I think I've got myself a bit mixed up here... I'm actually using the SMPS pretty much as it is... I think I'm actually using both secondaries, but I'm only using the +24v rail.... I was thinking there might be some way of making the -24v become +24v somehow, but I don't think it is possible without lots of hairloss and headaches... It works and that's fine for me.

 

[djQUAN]

you're already using both secondaries. the output rectifier is a bridge with center tap so using only one polarity output will only use two diodes in full wave rectification.

 

[luka]

Hi

Can you check for me, when you load only one seconday, you have one voltage there, but at that moment what is voltage on other side?

Well, I must admit, I wouldn't have a clue how to do this... I've measured ripple on the speaker outputs of my class-d module of around 330KHz, so this might be a bit too high frequency for the SMPS to run at...

You will/if at all, hear audiable freq. that shouldn't be there, even if no signal on input. You don't have to do it(sync...), until you know if you even have beating between smps and Damp(I don't have it with my smps and AMP2, or had with same smps and Damp that I made), then we will help you do that. Ripple at output @330kHz is small and it is fine, this is how amps of this nature work

 

[MikeHunt79]

Hi

Can you check for me, when you load only one seconday, you have one voltage there, but at that moment what is voltage on other side?

DJ Quan is right, I'm actually using both secondaries after all... I'm only using a single rail tho, despite this SMPS having +24 and -24v rails, I'm using only the 24v rail to power my amp module. If you like I can just load the +24v rail and see if the -24v sags also?

I will eventually use the -24v rail to power some op amps.

will/if at all, hear audiable freq. that shouldn't be there, even if no signal on input. You don't have to do it(sync...), until you know if you even have beating between smps and Damp(I don't have it with my smps and AMP2, or had with same smps and Damp that I made), then we will help you do that. Ripple at output @330kHz is small and it is fine, this is how amps of this nature work

My hearing drops off at around 16KHz, but when I was testing, I couldn't hear any hiss, hum, or high pitched noises... It sounded very good. BUT, I only tested for a short amount of time... Once it is in the car, I will be sure find any issues (if any).

 

[luka]

Hi

You don't have to test how would other sec react, no need it you will use both

If you couldn't hear now, then you won't hear at all, not now not ever, coz there is probably no beating present

More pic, hehe

 

[MikeHunt79]

Hi

You don't have to test how would other sec react, no need it you will use both

If you couldn't hear now, then you won't hear at all, not now not ever, coz there is probably no beating present

More pic, hehe

Cool, I've started to mount the amp module in the chassis anyway...

More pics. Last pic was 2 x 50w h4 bulbs (4.5A load), but here's 4 x 50w H4 bulbs (9A load), all croc clipped together, on a paper envelope!
[IMGDEAD]http://img248.imageshack.us/img248/2919/imgp1089nu3.jpg[/IMGDEAD]

[IMGDEAD]http://img404.imageshack.us/img404/8241/imgp1088jq1.jpg[/IMGDEAD]

Amp module just about squeezes into the chassis:

[IMGDEAD]http://img442.imageshack.us/img442/6417/imgp1100ej2.jpg[/IMGDEAD]

Just a bit more wiring and this thing should be ready to sing.

 

[luka]

Hi

Looks very good, but how will amp cool off, since it will be boxed inside?

 

[MikeHunt79]

Hi

Looks very good, but how will amp cool off, since it will be boxed inside?

I have bolted a block of Alu to the amp chassis, then I have bolted the amp module to the block of Alu, and it acts as a heatsink... The cooling will be much better than the original heatsink that came with the module.

I'll have to grab some more pics when I get the chance, but I'm just about to leave for work.