OK, since I seem to have thrown us in the direction of SMPS, I decided to try a few things this morning. I took a couple 63v, 4700uf caps and hooked them together to form a +- filter. I hooked them up too my SMPS300RE +-36V. I cycled on and off and wouldn't charge the caps. Then I hooked them up to this SMPS and it works. So I hooked it up to the FetZilla and still no problem. I was actually surprised at how good it sounds on +-25v rails. I'm not sure why this SMPS doesn't have a problem with the added caps and the connex does. I'm going to try going up in cap size and see if it can handle it.
Blessings, Terry
Blessings, Terry
Terry, i'm extremely obliged that you can test this. Could you please try this setup once more, but with a 10 ohms resistor in series between each rail output of the SMPS300RE and the caps? Did you use a single 4700 uF cap per rail or two or four? If you used two, could you try it with just two 6300 uF caps? Since the SMPS300RE already has 2000 uF of capacitance per rail we'd only need 7400 uF additional per rail(per channel) to reach the design spec of 9700 uF for the Fetzy, with the caveat that 2000uF of this capacitance is shared between left and right channels.
It works with the other SMPS probably because it has an arguably better-designed soft-start. I'm only guessing of course, but looking at the circuit board for the SMPS that works, i can see some power resistors close to the output. It could very well be it uses these to limit in-rush current. There are some small transistors close to the resistors, maybe it uses these to bypass the resistors after a while.
Terry, you have a scope right? Is it of the storage variety so that it may be possible somehow to "sample" the first few hundred milliseconds of the SMPS300RE when connected to the 10 ohms - 4700uF(per rail) filter?
I studied the datasheet of the controller IC used in the SMPS300RE, along with the low-res schematic Cristi provides in the PDF that goes with the SMPS300RE, and it seems altering the time constant dictating how long the SMPS takes to ramp up voltage to charge its caps is possible, but prohibitively difficult, since it's a single SMD capacitor in a hard-to-reach place. Maybe in the end a resistor or a Power NTC bypassed by a (solid-state) relay after a time is the way to go after all.
Bit of a bummer, since it would have been elegant to find a way to have more capacitance without the added complexity of a soft-start.
All the best,
Kris
It works with the other SMPS probably because it has an arguably better-designed soft-start. I'm only guessing of course, but looking at the circuit board for the SMPS that works, i can see some power resistors close to the output. It could very well be it uses these to limit in-rush current. There are some small transistors close to the resistors, maybe it uses these to bypass the resistors after a while.
Terry, you have a scope right? Is it of the storage variety so that it may be possible somehow to "sample" the first few hundred milliseconds of the SMPS300RE when connected to the 10 ohms - 4700uF(per rail) filter?
I studied the datasheet of the controller IC used in the SMPS300RE, along with the low-res schematic Cristi provides in the PDF that goes with the SMPS300RE, and it seems altering the time constant dictating how long the SMPS takes to ramp up voltage to charge its caps is possible, but prohibitively difficult, since it's a single SMD capacitor in a hard-to-reach place. Maybe in the end a resistor or a Power NTC bypassed by a (solid-state) relay after a time is the way to go after all.
Bit of a bummer, since it would have been elegant to find a way to have more capacitance without the added complexity of a soft-start.
All the best,
Kris
Last edited:
Hi Kris,
I used one 4700uf per rail. I don't have any 6300uf caps. Next size I have is 10,000uf. I just hooked it up with the 10W 10R resistors in line with each rail. The rails dropped to +-20V. I just noticed that I had a typo above. When talking about the SPMS300RE it should have said, "It cycled on and off and wouldn't charge the caps." The amp seems to play fine with the +-25V SMPS. It does begin to distort at a lower volume but sounds decent. I will try later creating a CRC pack and see if it can handle a start-up with that. Hopefully that SMPS has a shut down relay if over stressed so I don't burn something up.
Sorry, my scope is a Tektronics 475. An old beast. No digital info available. I barely know enough to look at some sine and square waves.
Blessings, Terry
I used one 4700uf per rail. I don't have any 6300uf caps. Next size I have is 10,000uf. I just hooked it up with the 10W 10R resistors in line with each rail. The rails dropped to +-20V. I just noticed that I had a typo above. When talking about the SPMS300RE it should have said, "It cycled on and off and wouldn't charge the caps." The amp seems to play fine with the +-25V SMPS. It does begin to distort at a lower volume but sounds decent. I will try later creating a CRC pack and see if it can handle a start-up with that. Hopefully that SMPS has a shut down relay if over stressed so I don't burn something up.
Sorry, my scope is a Tektronics 475. An old beast. No digital info available. I barely know enough to look at some sine and square waves.
Blessings, Terry
Hi Kris,
I used one 4700uf per rail. I don't have any 6300uf caps. Next size I have is 10,000uf. I just hooked it up with the 10W 10R resistors in line with each rail. The rails dropped to +-20V. I just noticed that I had a typo above. When talking about the SPMS300RE it should have said, "It cycled on and off and wouldn't charge the caps." The amp seems to play fine with the +-25V SMPS. It does begin to distort at a lower volume but sounds decent. I will try later creating a CRC pack and see if it can handle a start-up with that. Hopefully that SMPS has a shut down relay if over stressed so I don't burn something up.
Sorry, my scope is a Tektronics 475. An old beast. No digital info available. I barely know enough to look at some sine and square waves.
Blessings, Terry
Thanks Terry! So did you try the SMPS300RE with the 10 ohms or the other, 25V one? I was thinking of having you try the SMPS300RE - 10 ohms - 10.000uF cap (per rail).
And yeah the SMPS300RE when it detects overcurrent(over 5.1Amps) will shutdown, wait some time(1-2 secs?) and try again. It seems it's designed to just barely stay under that amperage when charging its built-in capacitors, and so it's very sensitive to caps at the output, or if the smps AndrewT refers to is indeed the 300RE, it seems it's fairly ill-suited to any Class AB amp, due to having problems with even fairly low quiescent currents.
The Tek 475, lauded by some to be the best audio scope
Wish i had a scope...
Yes, I tried it on the +-25V unit. I don't have the 300RE hooked up right now. I'll try to get to it soon.
I'm not sure how many amps it takes to kick off the 25V unit but it does shut down and reset. I accidentally shorted the rails and it kicked off and then came back on after a few seconds. Pretty tough little SMPS. I hope I can get it sorted out for audio use. Hard to beat $29 free shipping for a complete PSU.
I'm not sure how many amps it takes to kick off the 25V unit but it does shut down and reset. I accidentally shorted the rails and it kicked off and then came back on after a few seconds. Pretty tough little SMPS. I hope I can get it sorted out for audio use. Hard to beat $29 free shipping for a complete PSU.
I put together a small CRC to try with the 25v SMPS. I used 4, 4700uf caps and two 5W 1R resistors. The SMPS started fine connected to the caps so I hooked it up to the two amp modules and it still started up just fine. The only time it didn't was when I had everything hooked with a music signal being inputted. I turned down the input and then it started just fine.
The FetZilla sounds very good set up like this. I A/B'd it against my VSSA and they are very, very close until you get to high volume and then the FetZilla starts clipping sooner than the VSSA but to be fair, The VSSA is running on +-37V rails and a cap multiplier and FetZilla is +_24V after the CRC. I would be plenty happy to use it as is. It is also dead quiet with this setup and I don't even have it in a case and have wires running all over the place.
Blessings, Terry
The FetZilla sounds very good set up like this. I A/B'd it against my VSSA and they are very, very close until you get to high volume and then the FetZilla starts clipping sooner than the VSSA but to be fair, The VSSA is running on +-37V rails and a cap multiplier and FetZilla is +_24V after the CRC. I would be plenty happy to use it as is. It is also dead quiet with this setup and I don't even have it in a case and have wires running all over the place.
Blessings, Terry
Attachments
OK, one more thing to add. I hooked up the 300RE using the same little CRC and inserted CL-90 thermistors between the SMPS and the CRC filters. Upon first fire up it worked fine. It drops about 2.1v per rail measuring from SMPS output to tab on the boards. This is driving both boards with music playing. The issue arises when you shut down the amp and then try to restart it. The SMPS starts its cycling on and off. You have to let the amp sit for while so the CL-90s can cool down. Then it will start again. So this tells me you could use a soft start circuit between the SMPS300RE and the filter bank and have it short out after a few seconds and you should be fine.
Blessings, Terry
Blessings, Terry
Hi Terry,
Would you kindly give a simple sketch on this with the soft start.
Thanks
Sorry, I'm not sure how to make the soft start work on the DC side of the SMPS. I'm sure there must be a way. Probably one of the engineers here can draw it up. A 10ohm resistor would probably be enough but you would want to short it after the caps were charged so you didn't lose all that current. The soft start won't work in front of the SMPS because it is designed to use a wide range of input voltage and still produce it's given value.
And the resistor should be a wirewound type able to withstand very high current for a very short time. Apparently not all wirewounds are created equal. Some more investigation is in order. And we also need to take into consideration the switch-off behaviour. I.e the cap providing the time constant for the relay/solid-state solution which shorts the wirewound/thermistor should be able to keep it shorted for long enough to drain both the caps of the SMPS as well as the reservoir caps of the Fetzilla, lest we are likely to shorten the life of the bypass element, especially if it's a relay.
I'll try to sketch something up as soon as i can, i have an idea but need to research components.
All the best,
Kris0603
I'll try to sketch something up as soon as i can, i have an idea but need to research components.
All the best,
Kris0603
Last edited:
Hello everyone,
I've started to build the Fetzilla this week-end and saw that I may have made a mistake in my order. Well, there were several mistakes in the order as I was using the .doc from swordfishy BOM and there are mistakes in that and missing parts and all but meanman1964 was kind enough to catch most of them for me. That being said, I ordered PRP 1/4W resistors instead of 1/2W. Is this a really bad mistake? Do I need to change all resistors?
Thanks.
Bertrand
I've started to build the Fetzilla this week-end and saw that I may have made a mistake in my order. Well, there were several mistakes in the order as I was using the .doc from swordfishy BOM and there are mistakes in that and missing parts and all but meanman1964 was kind enough to catch most of them for me. That being said, I ordered PRP 1/4W resistors instead of 1/2W. Is this a really bad mistake? Do I need to change all resistors?
Thanks.
Bertrand
No, you don't need to change them all.
Calculate the power dissipation in each at quiescent conditions.
If any are below 25% of Pmax, i.e. <60mW then they are safe to use.
If any are between 26% and 50% of Pmax then they will run warm. This reduces their life slightly. It will also change their value slightly.
Both these effects will not stop the amplifier working for the first week or two.
Any that are above 50% of Pmax should be changed.
You can use two 1k0 250mW in parallel to give an equivalent 500r 500mW. One above the PCB and one below the PCB. Or both above the PCB, one above the other.
Similarly two 240r 250mW in series is equivalent to 480r 500mW. Not as neat to solder into a PCB, but it works until a suitable replacement arrives.
This saves throwing away what you have.
Calculate the power dissipation in each at quiescent conditions.
If any are below 25% of Pmax, i.e. <60mW then they are safe to use.
If any are between 26% and 50% of Pmax then they will run warm. This reduces their life slightly. It will also change their value slightly.
Both these effects will not stop the amplifier working for the first week or two.
Any that are above 50% of Pmax should be changed.
You can use two 1k0 250mW in parallel to give an equivalent 500r 500mW. One above the PCB and one below the PCB. Or both above the PCB, one above the other.
Similarly two 240r 250mW in series is equivalent to 480r 500mW. Not as neat to solder into a PCB, but it works until a suitable replacement arrives.
This saves throwing away what you have.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.