The recent donation of the APT 105 AC Pwr Sply made this project really easy. Measurements were made at 115, 120, and 125 vac. The pwr sply can give you the primary current or watts. The transformers involved were a standard E core rated at 40v 10a and a toroid rated at 2x 25v 10a.
Setup for the E core transformer. Efficiency: ~90 - 92%
Setup for the toroid transformer. Efficiency: ~91 - 96%
So what projects have you done in the ELab lately?
Just a quick note about the power supply that was used in the above experiment. It is an APT 105 Power Converter. It can provide an isolated output voltage from 0 to 300 VAC, either 50 Hz or 60 Hz, and up to 500 VA. At 115 VAC it can deliver 4.6 Amps, at 220 VAC it can deliver 2.3 Amps. It has digital displays for Voltage, Current, and Power. The equipment also has a universal output jack that can accept all international power plugs. Let’s find another good use for all the fine instrumentation in the Electronics Lab.
All we need is a programmable load and we could do some power supply testing. Combine with the NI Virtual Bench and it could be automated. I have a custom 24vac to 5v supply i designed and still plan to test it. just got sidetracked with all the software development aspects of the project.
How about a B&K 8610?
Does the APT 105 have external control capability ?
If you want to do automated testing, it would need some interface that LabVIEW can talk to.
Did a quick look at the website. Doesn’t look like it has any com ports of any kind. Full manual mode.
The manual Art linked shows that the supply supports GPIB and RS232
Edit: I did this yesterday with an Agilent e3631a using Python and PyVisa over RS232
Dallas Makerspace Show and Tell - March 2017
There are no external interface ports on the APT 105. Art is right, just manual mode.
Maybe “automated testing” was a bit of a reach. Could still do automated data collection, but would need to manually turn the voltage knob on the APT 105 for each data point.
If you parallel a step down transformer or voltage divider you could measure the output voltage. Add a current transformer or hall sensor and you can measure current. You could measure all you need for each voltage / frequency that way. Just have to change the knob to move to a new setting.
There are times when “quick and dirty old school” will get the job done faster. This was one of those times. For the E core we did secondary loads of 4 and 8 ohms for the 3 input voltages resulting in only 6 data sets. For Walter’s toroid we did several loads for each input voltage. The 2 secondaries were connected in parallel. I think Walter created a spreadsheet in Libre Ofc to do the calculations. The primary purpose of the set up was to determine the efficiency at partial and (near) full load. Interesting observation - the efficiency of the toroid went down and then back up.
All that said, this would be a good basic design exercise to automate. More data points and graphing of the data would be interesting, but gross overkill for what I sought.
One of the long term goals I have for the ELab is to update/ obtain gear that can be used in an automated test set up using LabView or ?. For example do a swept freq input to a audio/ rf/ fltr ckt to generate response curves. Anybody who has done this manually will tell you it can quickly get laborious/ tedious.
The BK 8601 would easily satisfy any hobbiest the aspirations I might have.
I agree with Art. The newer generation, those folks born with a mouse in their hand, always assume that a computerized test is the only way to go. Automatically collecting more data into a spreadsheet using LabVIEW or whatever application is not always the answer, it’s just more data. If one understands how the circuit under test is supposed to operate, than a few data points is all one really needs to know to complete a specific test. The fact that Art and Walter completed their entire test; 1) agreed on a test plan, 2) built the test circuit, 3) used available instrumentation, 4) collected their data, 5) verified the data was correct, 6) analyzed the results, 7) even took a few pictures of their test setup, and 8) published their results, all in a mere hour or two is the real success story here.
Lol. I wasn’t born with a mouse in my hand and the web browser didn’t exist when I was in college. I have managed a small test facility with a dozen employees. Automation is not about doing something that seems impossible otherwise. It’s also not always about doing it faster. Many times It is about building repeatable processes that can be accomplished by less skilled operators. Power supply load testing for a bunch of makers that rarely need the skill seemed like a good candidate. The reality is that with the right programmable load and an oscope, I can do all the load testing I need; I think transient response is important for regulated supplies. The NI comment was just throwing out cool things we could do with existing tools.
LabVIEW is a great tool, I do not disagree with that. Plus, LabVIEW is available at the DMS Electronics Lab. But, it does not have the interfaces required to connect to the equipment that was used in the Transformer Efficiency Test. My comment was based on the use of readily available test equipment and that everything was accomplished in a few hours.
