Aleph-X builder's thread.

Yea, I was waiting for someone to recognize my idea above..... I work with alot of 3 phase rectifiers in the Motion Picture industry, some capable of 400 amps continous @ less that 5% ripple. Ripple is alot easier to deal with and inductors are physically much smaller to boot. Unfortunately not too many homes have 3 phase capabilities in them.

Lets see 400 amps at 30 volt rails....that would be one hell of a huge X amp and could probably drive loads as lows as .1 ohm at full power! I think the heat exchanger would be a tad bit larger than what you were thinking of though.......:bigeyes:

Mark
 
Update on the biasing resistors...
Out of 10 .25 ohm Dale 1%ers I got 8 that were .23 ohms, two @ .24 ohm. Needless to say I used the all of .23's as the stock design calls for .22 ohms. Unfortunately I saw CheffDeGaars posting after I had already ordered the components...but what the heck...I'm only building this one channel for now to understand how it all works... Now on to FET matching... then hook up the supply and see what I get.......

Mark
 
nead help, how to macth resistors

Hi all...
I need help how to match this low value resistors...I have 0.5 Ohm mills resistors and I try to match them but the curent and voltage drop was to little to measure...
I have regulated 30V 940mA from my printer power supply...I've got two multimiters, one fluke 87III and other some cheap no name , but it's working...will this work...
Any suggestion how to mach them???
How do you mach this small value resistors???

Thanks
Josip
 
We may not have 3 phase, but most car altinators can put out 100 amps and up. the voltage is regulated by the feild current x rpm. I think a 3 hp electric motor turning a altinator 4000 rpm could get you +- 30 volt rails at 30 amps at 800 hz or so. wouldn't take much capacitance to smooth that out nicely. And it would be cheaper than 2 toriods(or about the same price)
 
Let me try to run the numbers in an understandable fashion.

Lets say you want 64W RMS. That's 128 watts peak. For 8 ohms, that's Sqr(128/8) = 4 Amps , at 8 ohms, thats 32 volts.

The AX is a bridged amp, so each half needs to swing +- 16 volts at 4 amps.

Giving 4 volts to be lost in the current source at maximum swing, that means 20 volt rails, each side biased at around 3 amps. The Aleph current source will peak 4.5A at 50% gain (this is the part where I'm a little unsure about the math) Every one says more bias sound better. If your speakers dip below 8 ohms, bias to their minimum impedance at minimum.

So you have 4 banks of output devices burning 60W each. You should try to keep dissapation per device around 25 watts each to be safe. 2 devices per bank is 30W each ( a little close to the edge if your sinks aren't large enough, but should work ) or 3 devices at 20W each.

Refigure your source resistors so they drop .5V +- for the current in each fet. At 2 devices, 1.5A each is .33 ohm, at 3 devices, 1A each, .5 ohm. You may also need to adjust the current source for the diff pair so you don't burn the zener.

Going through these steps will give you a better understanding of the relations than plugging numbers into the spredsheet thats somewhere else in the forum.

All corrections are welcomed.
 
Ex-Moderator
Joined 2002
To those getting their new AX boards or if your finally going to use the boards from last years sale, if you need parts, read on.

If you're a Cardas fan, the 2nd annual Cardas bulk purchase is well under way. If you're interested see this thread in the trading Post Forum and the WIKI Page for details.

I have started to send order confirmation emails to those that already placed their orders. I haven't processed them all yet but yours is coming soon.

I will be accepting requests for the price list up until Wednesday, the 25th at 12 midnight EST.

PS: I will delete this post after Wednesday.
 
Re: nead help, how to macth resistors

yoke said:
Hi all...
I need help how to match this low value resistors...I have 0.5 Ohm mills resistors and I try to match them but the curent and voltage drop was to little to measure...
I have regulated 30V 940mA from my printer power supply...I've got two multimiters, one fluke 87III and other some cheap no name , but it's working...will this work...
Any suggestion how to mach them???
How do you mach this small value resistors???

Thanks
Josip


:cannotbe: :cannotbe: :cannotbe: any one???:bawling:
 
Re: nead help, how to macth resistors

yoke said:
Hi all...
I need help how to match this low value resistors...I have 0.5 Ohm mills resistors and I try to match them but the curent and voltage drop was to little to measure...
I have regulated 30V 940mA from my printer power supply...I've got two multimiters, one fluke 87III and other some cheap no name , but it's working...will this work...
Any suggestion how to mach them???
How do you mach this small value resistors???

Thanks
Josip

Just a silly idea I get :eek: :) . Not tested it, but may be worth a try :confused: Just take an opamp, preferably a fet input one, and make a non inverting amp just like the schematic attached. Hook the right supply and a AF generator, around 200Hz in frequency. Amplitude's up to you. Connect a DMM in AC voltage mode at output. Make sure the output amplitude of the generator is stable and don't touch it anymore Most of DMMs have a sufficient bandwidth in AC measurements to provide good reading at 200Hz. Just change your resistors, and note the output voltage value. Same resistors will give same readings... The feedback R on the schematic gives a voltage gain of 21 for a 0.5R "unknown" resistor. Just change it to suit your needs and your DMM sensivity.

Should work with a DC source, but the stability of supply and components should affect the readings.

Too silly to be good, but who knows... ;)
May be this helps :)
 

Attachments

  • image2.gif
    image2.gif
    2.8 KB · Views: 6,563
Thanks Cheff...
Now this maid be good idea for someone who is not total newbie as I’ am...and making circuit like this would take additional costs, and I don’t have AF generator ...I was hopeing for some simple circuit with two resistors or something...
Which everybody can use...I believe that here is many newbies like me who would like and have nead to match resistors...

Thanks once again Cheff ...
Any other sugestions ???

:smash:
 
Matching Resistors

Ok, there seems to be need for an explanation on how to match and more importantly MEASURE low resistance resistors.

There are ways to measure low resistances, and one of the most efficient methods is what is called the 4-wire method. In short this method uses a precision current source and a precision voltmeter.

When measuring the current source is set up to run a precisely known current through the resistance, and the voltmeter is used to measure the voltages that builds up over the different parts of the circuit.

Normal multimeter doesn't go as low in the resistance measurements. This is because the lower values will be inaccurate due to losses in wires, connections and so on. The only way to get around these problems is to use separate circuits for supplying the current that is needed to measure the resistance and the voltage measurement. This is after all the method used, but at normal values these problems doesn't play much of a role.

To give an example on how to do this: To make calculations easy let’s say we use a 1A current source. Now to measure the resistance on the parts all we need to do is to measure the voltage over the part. Now if we take a 0.33 ohm resistor as an example 0.3300V should be ideally measured across the resistors. This range is perfectly measurable with a standard multimeter with fairly good accuracy. With a 4 1/2 digit multimeter you will be able to measure resistance with better than milli-ohm accuracy! This means you will even be able to measure the resistances in the contact points! And all you have to do to measure them is to measure the voltage across them and not the resistors. Remember what you are measuring are voltages, and there are no problems with losses when measuring them, and you are using almost infinitely larger inner resistance in your multimeter so your measurements will not affect the measurement.

If you want to simplify it you can dispense the dead accurate current source if you hook in a current meter in series with the measurement. That way all you have to do is use ohm's law to calculate the resistances.

The downside of this measurement method is the high currents involved. The heat losses in the components will soon be a problem if the measurements aren’t done on power resistors, especially when the resistances are increasing! If you run the same 1A through a 10 ohm 0.6W resistor you'll fry it in seconds! So this method requires knowledge and thinking before using!
Elector has at least one really good article on building a really good low resistance measurement box. This also uses the low duty cycle method to avoid the heating problems. The current is on 1/100’Th of the time, and left off the rest. This way the heat in the measured device is only 1/100’Th of what it would be without. I have it somewhere but can’t find it now (and it is in Swedish...). If interested you should be able to find it if you do a search on their site.

But there are even more advantages, and that is the super matching possibilities! Let’s say you have 30 resistors you want to match. All you have to do is to connect them all in series with a suitable voltage source and a current-meter. If we take the 0.33 ohms resistors from the example above you need aprox. 0.33V per resistor (if you want to test at 1A), and with 0.33V X 30 = 10V. If you want to test them at a different current (for example the actual current you are going to use in your amp) all you have to do is calculate the corresponding voltage. You have to see to it that they have the same cooling possibilities they have in the amp! And also possibly a heating period to stabilize in temperature.

Now all you do is measure the voltage across each of the resistors. And remember that the connections have resistances too, and will be taken in to the measurement if you don’t remember to place the probe at the right spot! All you have to do now it to match the ones with closest voltages! And since they share the current it will be perfectly equal in each resistor. So the voltage over each resistor is a perfect indication on the resistance.

Hope this helps! Now you should be able to match the resistors even better than the transistors!;) And this can even be done both easy and without expensive equipment! :smash:

Anders
 
Question about Q12a/b

Hi All!
The boards just arrived yesterday (they are really great, thanks Chad) and I'm starting to look them in deep.

I have questions about Q12a\b option.
Is it right that just one of those regulators mast be used? I know is a silly question, but my english can't give me the absolute safeness for this.

In addition, do I have to make the same choice for j1a\b? Is this choice related to the previous (q12a = j1a and q12b = j1b) ?

Thank you for your support.
Bye!
 
Do you guys have any suggestions for starting values of C6, C7, and C8?
C7, and C8 are optional,
But c6 doesnt say it is... (Ref BOM via Hifizen) :confused:

Is it true that transformers are suggested compared to SMPS's due to hum and line noise being carried into DC?

And what ar eyou guys going for heatsinks? it'd be nice to have a group order of correct heat sinks

I'm thinking about getting a 80$ set of pneumatic tools (for use with air compressors), that has a drill, a grinder (hand), and a saw... does it sound like a good deal? (i have a craftsman air compressor) (TOols are for the aleph-x's chassis :))

edit:
For R49, should this be jumpered? or not? or should an actual 0 ohm (burn-up protection?) resistor be used? Any starting value ideas for this?
For R48, any idea-rs for a startin value? (Optional resistor to reduce power dissipation in q6*)
Thanks in advance :)
 
Value of R11 and R33

While we are on the subject on what values resistors should be.... What is the correct value for R11 and R33? In this post the values of these resistors seems to make some problems when using the recommended 47k5! When replacing with 4k75 as some of the schematics floating around seems to use the problem is solved.... Part of the problem as I see it is what values should VR1 and VR3 be set in when the initial testing is done. If they are set at maximum values the total resistance of R11 and VR1 is 147k5. Will this mean full current through the amp?

Just to clarify this, how should the variable resistances be set before testing, and is it safe to use the recommended 47k5 values (or will it possibly blow up the transistors like it did to Blitz)??

Anders