I assembled the power supply with Jan's recommended resistor changes and have the same MBR1045 heat issues as others in this thread.
I am a bit skeptical that the MBR1045 is at fault when the MBR1060 or MUR820G work fine.
The 47nF snubber seems suspicious to me. I'm not an expert but typical values I've seen in this position are below 1nF.
I will experiment with removing 2 caps and leaving the other 2 and check for a temperature difference.
Related, in LA #5 (same issue as this preamp), there is the Rectifier snubbing article by MJ suggesting that bypassing each diode is a faulty approach. He recommends a resistor-capacitor per transformer secondary. Another option is to go with the Quasimodo C-R-C. Perhaps these changes can be incorporated for future revisions of the PSU PCB
I am a bit skeptical that the MBR1045 is at fault when the MBR1060 or MUR820G work fine.
The 47nF snubber seems suspicious to me. I'm not an expert but typical values I've seen in this position are below 1nF.
I will experiment with removing 2 caps and leaving the other 2 and check for a temperature difference.
Related, in LA #5 (same issue as this preamp), there is the Rectifier snubbing article by MJ suggesting that bypassing each diode is a faulty approach. He recommends a resistor-capacitor per transformer secondary. Another option is to go with the Quasimodo C-R-C. Perhaps these changes can be incorporated for future revisions of the PSU PCB
I was going to use one of those two listed to sidestep that issue, but ordered the FES8GT-E3/45 since it was in the BOM. I haven't received them yet to verify their operation. Anyone else try these?
I removed the 47nF capacitor and the MBR1045 was still very hot, so that does not seem to be the issue.
There were a few MUR820's in my parts bin so I put them instead of the MBR1045 and they remained cool to the touch.
I don't know what the problem is but maybe best to make a BOM change since this has been an issue for a few builders.
Edit: Is it possible they are simply underrated? Do they see twice the peak voltage or is it different in this configuration? 18V*1.414*2 = ~50V, and the MBR1045 is rated at 45V.
There were a few MUR820's in my parts bin so I put them instead of the MBR1045 and they remained cool to the touch.
I don't know what the problem is but maybe best to make a BOM change since this has been an issue for a few builders.
Edit: Is it possible they are simply underrated? Do they see twice the peak voltage or is it different in this configuration? 18V*1.414*2 = ~50V, and the MBR1045 is rated at 45V.
Last edited:
Tested the PSU again tonight. The transformer has 115V primaries while my mains are around 123V, resulting in 23VAC secondaries when lightly loaded.
I hooked up the scope, and with 10V/div, there is just over 60V across the diode.
I'm still learning... does that entire 60V cycle apply to the Vrrm rating of 45V, or just part of it?
The MBR1060 is rated at 60V (maybe that's why it did a bit better), the MUR820 at 200V and the FES8GT-E3/45 at 400V, so I think any of the latter two would be preferable.
That aside, the PSU is easily configurable to 17V and is rock steady once adjusted.
I hooked up the scope, and with 10V/div, there is just over 60V across the diode.
I'm still learning... does that entire 60V cycle apply to the Vrrm rating of 45V, or just part of it?
The MBR1060 is rated at 60V (maybe that's why it did a bit better), the MUR820 at 200V and the FES8GT-E3/45 at 400V, so I think any of the latter two would be preferable.
That aside, the PSU is easily configurable to 17V and is rock steady once adjusted.
Attachments
Parts should be applied with "margin" with respect to most of their parameters. A diode rectifier for example, should be specified to be used at no more than 70% of its rating for what would be called an "Enterprise" class products. "Consumer" class products might go with a 90% de-rating. As a DIYer you get to decide what your reliability goals are. If you are good with "Are you feeling lucky today, punk?" reliability, go ahead and use parts at their full rating. In this case, the diode is being stressed to the full voltage shown in the scope shot.
Mouser has the 4PD, search for 688-SPUN194700. I bought the RCA jacks at Newark. Search for 26M2040. Newark is out of the white ones, so just use two of the red ones and change the color of one using a marker.
Fantastic. Let's do this then. Let's see if someone else will chip in with the microcontroller code. I too was looking at Arduinos and I too thought that the Nano was perfect for this. I vote that we set up the Arduino to read the codes of the Apple Remote -- it looks so sexy, and it's available all over.
If you are only looking for remote control volume then follow the arduino code here:
https://github.com/gkiranps/gArduinoRemoteThe code uses most common remotes and you can modify the code according to the remote you use. You can use multiple digitals pins or use a h-bridge to drive the motor.
Now if you want to use arduino to control source selection and mute, its a bit more involved but doable.
This is really great, thanks.
Can you point me to any details, any schematic, for controlling the motor? Will this part work? https://www.mouser.com/datasheet/2/115/ZXBM5210-335426.pdf
A YouTube video I saw used the chip https://www.mouser.com/datasheet/2/389/cd00000059-1795435.pdf which is physically larger and thrice as expensive. A web page (URL here: https://www.electronics-lab.com/project/motorized-slide-potentiometer-driver/) too refers to the same chip. I guess there will be cheaper and simpler alternatives? I guess the motor drive can also be done using two pairs of BD139/BD140 transistors, but I'm not smart enough to design the circuit on my own. I believe a few additional parts need to be added for protecting the transistors against over-current or reverse EMF from the motor coil, etc.
Last edited:
I found this page which controls the motor using four FETs which are directly driven by the uC pins: https://www.ianjohnston.com/index.p...amplifier-motorized-volume-control-conversion
This article uses BJT to drive the motor: http://electronics-diy.com/electronic_schematic.php?id=691 Of course, the digital circuit which drives the transistors can be replaced with an Arduino etc.
I don't know enough to evaluate how good or bad these are.
This article uses BJT to drive the motor: http://electronics-diy.com/electronic_schematic.php?id=691 Of course, the digital circuit which drives the transistors can be replaced with an Arduino etc.
I don't know enough to evaluate how good or bad these are.
The easiest would be this part - L9110H H-Bridge Motor Driver for DC Motors - 8 DIP - 2.5V-12V 800mA: https://www.mouser.com/ProductDetail/485-4489.
You could alternatively build a discrete h-bridge using npn and pnp transistors as shown at ESP site (figure 2): https://sound-au.com/project110.htm
I have used both and they work very well.
Hard to beat an integrated 8 pin chip versus a discreet circuit for an H-bridge motor driver. Note that the "classic" problem with an H-bridge circuit is you must insure that there is no overlap with the top and bottom driver on a given side of the circuit being on simultaneously. Not too difficult to achieve with such a slow speed circuit, but why bother?
Great, thanks, I had missed the ESP page. But the L9110H chip: it's the strangest chip I've encountered. Is the manufacturer actually Adafruit? The datasheet seems to be a bad translation from some other language. But the chip is simple enough and inexpensive, so I guess it's usable.
The chip is not strange. Yes you could almost carve one out in your bedroom. However there are people in oriental alleys with 1999 machinery looking for very-simple chip designs they can make cheaper than the Big Boys. And there is a vast Chinese chip-market which does not need docs in English.
Yes, AdaFruit is very USA. But LadyAda is always looking for cool electronics. I would guess she saw this L9110H in products, has agents who know how to do business in China, and bought a bunch. And while she does sell out of her own store, lately she realized that her talent should go to new ideas and products not boxing. And she now has enough clever products that the Distributors are pleased to do her stocking and shipping.
While Mouser would normally put the maker's trademark on their listing, nobody shopping at Mouser knows who this foundry is; the foundry may not be able to take questions in non-Chinese languages; also there are several companies with similar names; also this one may have changed name or gone out of business. The Adafruit brand is good.