Does a hobbyist need a bench power supply?

[cybergibbons]

Thanks for the kind comments!

The other power supply, the switchmode PSP-2010, I took apart when I first got it (in the true spirit on EEVblog I guess...).

The construction quality of this is fairly good again. They seem to have added a sheet of plastic to duct airflow over the choke and power resistor, but outside of that they seem to have got this right first time.

There's a good number of ferrite beads, threadlock on most screws, soldering is neat.




In use, I'm more than happy with the power supply. The controls are digital, which means I can dial in x volts with a limit of y amps. It also has limits which you can set, so that if you are testing a circuit that has a max vcc of 12v, you can cap it there, and not worry about spinning the dial too far. The power limit is also an interesting feature, but I have yet to use it.

RS232 is a good addition. You can both control and monitor the supply using this - the command set is very simple. I'm contemplating adding a cutout and FTDI board so that I can control it with USB. Currently I'm using USB->Serial adapter from about 10 years ago and it's crap.

It's a useful supply for powerful projects - motor controllers, RF amps. Also really useful to use as a charger for different battery types as you can monitor and change the charging profile from a PC - I've not written anything to do this yet as the USB->Serial adapter fails all of the time!

[saturation]

You're very welcome.  Thanks again for the dissection. 

Lots of ferrite beads ~ lots of EMI, as expected from switching PSU, including much smaller, and cheaper type, laminated core transformers.  Construction is as you say, very well done. 

Have you tried looking at its ripple under load, comparing the switched PSU versus the analog one you bought, also at near maximum load?  I'm interested in whether the published performance spec is near its real world measured spec.

Thanks for the kind comments!

The other power supply, the switchmode PSP-2010, I took apart when I first got it (in the true spirit on EEVblog I guess...).

The construction quality of this is fairly good again. They seem to have added a sheet of plastic to duct airflow over the choke and power resistor, but outside of that they seem to have got this right first time.

There's a good number of ferrite beads, threadlock on most screws, soldering is neat.

In use, I'm more than happy with the power supply. The controls are digital, which means I can dial in x volts with a limit of y amps. It also has limits which you can set, so that if you are testing a circuit that has a max vcc of 12v, you can cap it there, and not worry about spinning the dial too far. The power limit is also an interesting feature, but I have yet to use it.

RS232 is a good addition. You can both control and monitor the supply using this - the command set is very simple. I'm contemplating adding a cutout and FTDI board so that I can control it with USB. Currently I'm using USB->Serial adapter from about 10 years ago and it's crap.

It's a useful supply for powerful projects - motor controllers, RF amps. Also really useful to use as a charger for different battery types as you can monitor and change the charging profile from a PC - I've not written anything to do this yet as the USB->Serial adapter fails all of the time!

[cybergibbons]

I'm going to do some tests and post them - but I do need some consistent loads for both of them - 20V at 10A means I need to find something that will dissipate 200W of heat!

[saturation]

Yes, only if easy enough for you to do.  You mentioned you had an ~ 32V/3A load somewhere, and the 4303 has a load regulation spec at 3A of ~0.01%.  The PSP 2010 spec is written independent of load, so you could compare it at open circuit and at 3A too.

https://www.eevblog.com/forum/index.php?topic=683.msg8431#msg8431

I'm going to do some tests and post them - but I do need some consistent loads for both of them - 20V at 10A means I need to find something that will dissipate 200W of heat!

[Zed]

When you build your PSU use digital for the voltage display and have an analogue display for the current (Moving coil meter or LED Bargraph). 7 Segment LED/LCD displays for current measurement without a bargraph display are almost totally useless even though you see them often on manufactured PSUs.

DO NOT PAY GOOD MONEY FOR A MANUFACTURED PSU! They are easy to build yourself.

[cybergibbons]

There is no way I could have matched the value of the 4 channel power supply I got - just a quick calculation would put the case, analogue panel meters, terminals, transformer and fan over cost. I should imagine the design has a few more features than the basic supplies I often build.

Does anyone actually really trust panel meters when it counts? I always double check the voltage with a meter, and if the current matters to me, I short it to put it into CC mode and confirm that it is right. I'd rarely use it for monitoring.

[desolatordan]

When you build your PSU use digital for the voltage display and have an analogue display for the current (Moving coil meter or LED Bargraph). 7 Segment LED/LCD displays for current measurement without a bargraph display are almost totally useless even though you see them often on manufactured PSUs.

Please extrapolate. Do you prefer an analog display because you don't need the precision in constant current mode? I use the constant current mode for a number of things and require it to be just as precise as in constant voltage mode.

[alm]

Analog meters have a much higher bandwidth, and will show you fluctuating current draw much better. It's also easy to see a needle go to zero or full-scale from the corner of your eyes.

[cybergibbons]

Of course, you also get the tell-tale ticking as the needle slams into the stops!

[alm]

True. I find relays clicking (when a DMM changes ranges) also quite useful for this purpose .

[Zed]

When you build your PSU use digital for the voltage display and have an analogue display for the current (Moving coil meter or LED Bargraph). 7 Segment LED/LCD displays for current measurement without a bargraph display are almost totally useless even though you see them often on manufactured PSUs.

Please extrapolate. Do you prefer an analog display because you don't need the precision in constant current mode? I use the constant current mode for a number of things and require it to be just as precise as in constant voltage mode.

When you are supplying power to something the voltage will remain pretty constant (unless there is a fault) but the current will vary proportional to the load. If the load varies then so will the current. With an analogue current meter you can see instantaneous changes in current but with a digital current meter you will see numbers all over the place if the load is not constant.
Digital Voltage Meter. Analogue Current Meter. This is THE LAW!

[NiHaoMike]

When you are supplying power to something the voltage will remain pretty constant (unless there is a fault) but the current will vary proportional to the load. If the load varies then so will the current. With an analogue current meter you can see instantaneous changes in current but with a digital current meter you will see numbers all over the place if the load is not constant.
Digital Voltage Meter. Analogue Current Meter. This is THE LAW!

A row of LEDs can react even faster than an analog meter and be cheaper to implement.

Or how about have a color LCD for the front panel display and allow the user to select the display format using software? Maybe even have a basic "oscilloscope" that can resolve voltage and current variations up to something like 20kHz or so.

[Hypernova]

When you are supplying power to something the voltage will remain pretty constant (unless there is a fault) but the current will vary proportional to the load. If the load varies then so will the current. With an analogue current meter you can see instantaneous changes in current but with a digital current meter you will see numbers all over the place if the load is not constant.
Digital Voltage Meter. Analogue Current Meter. This is THE LAW!

A row of LEDs can react even faster than an analog meter and be cheaper to implement.

Or how about have a color LCD for the front panel display and allow the user to select the display format using software? Maybe even have a basic "oscilloscope" that can resolve voltage and current variations up to something like 20kHz or so.

A display like that will add a non insignificant chunk of coin to the price that most ppl aren't will to pay for. You are better off making your own small amp meter and have it sit between the terminal and your load.

[DaveB]

Along these lines does someone have a general opinion on the new BK 9110? http://www.bkprecision.com/products/model/9110/100w-multi-range-60v-5a-dc-power-supply.html

[saturation]

A very interesting unique device, roughly its limited by "constant" power, so it delivers more amps at lower voltage and lower amps at higher voltage to the limits of 5A or 60V, whichever is reached first.  Its programmable with memories, so its can memorize for quick recall a list of oft used combinations.

The specs are good, and the price range is what can be expected of B&K.

The 2 issues are, do you need programmability given its ~ $300 price and what do competitors offer?

The 'constant' power approach is unique to this B&K.   A traditional power supply state amp and voltage limits, e.g. all linears by Maxtech for example, to 30V is about $80, $50 for 15V both at 3A, with similar specs for power regulation but without the programmable conveniences.

http://www.mastechpowersupply.com/dc-power-supply/linear-power-supply/mastech-regulated-variable-dc-power-supply-hy6003d-60v-3a-new-model/prod_119.html

Along these lines does someone have a general opinion on the new BK 9110? http://www.bkprecision.com/products/model/9110/100w-multi-range-60v-5a-dc-power-supply.html