RoadTest: Programmable DC Electronic Loads with the BK8600
Author: Kilohercas
Creation date:
Evaluation Type: Independent Products
Did you receive all parts the manufacturer stated would be included in the package?: True
What other parts do you consider comparable to this product?: None
What were the biggest problems encountered?: Battery capacity test was hard to setup without using manual. Manual was good, and after spending 30s i found how to do it.
Detailed Review:
Hello.
Today we are going to look at BK precision BK 8600 DC electronic load. After a long delivery time i finally have a time to play with this load. As a name suggest, DC electronic load is a way of loading your power supply or any other device that produce power, in order to check how it performs. Where are many reasons why we need DC electronic load. For me, most important is to see how my power supply performs over various conditions, like drawing current, drawing power and how it compares with parameters of my power supply like switching frequency, input voltage and so on. By having DC load i can quickly create scenarios for my power supply, and select appropriate input voltage or switching frequency, in order to get maximum efficiency and minimal thermal dissipation. I will show the demo how my power supply performs over various conditions as well as try to maximize efficiency.
First lets look at the product range and features set. Quickly by going into manual, we can found features set, and from my point of view, they are more than i need. So good job BK Precision !
| Model | Power | Voltage | Current | Cost |
| 8600 | 150 W | 120 V | 30 A | $1,050 |
| 8601 | 250 W | 120 V | 60 A | $1,295 |
| 8602 | 200 W | 500 V | 15 A | $1,345 |
| 8610 | 750 W | 120 V | 120 A | $2,450 |
| 8612 | 750 W | 500 V | 30 A | $2,950 |
| 8614 | 1500 W | 120 V | 240 A | $3,625 |
| 8616 | 1200 W | 500 V | 60 A | $4,150 |
Many of these features are self explanatory, like CC, CV, CW, CR and so on. But others are more interesting.
List Operation
List mode lets you generate complex sequences of input changes with rapid, precise timing. This is useful when running test sequences with a minimum amount of overhead.The parameters of List operation include the name, number of steps (2-84), step width time (20us-3600s), and every steps’ set value and slew rate. The list file can be saved in non-volatile memory where it can be quickly recalled. Users can edit up to 7 groups of List files in CC mode only. In List operation mode, the electronic load begins to enable the List operation when it receives the trigger signal and will continue until the List operation is completed or the instrument receives another trigger signal. This mode can be used to test step response and other parameters that can help you to figure out loop compensation in your feedback circuitry based on your application. It is always trade-off between noise, overshoot, undershoot and so on. So quick test. Generated table for 10 steps each spaced at 1s with incremental current step, and in few minutes using external current monitor and MDO3104 scope get this waveform. Sorry could not get precise current representation of 3A/1V (0-10V = 0-30A) since scope does not have division factor i need. But result is very clear. By selecting low list duration, you can get LDO or DC/DC converter step response. This is so simple.
So this is very cool and simple to use. Just need to read user manual and you are good to go ! yellow trace is current monitor, and blue one is voltage on the DUT.
eGaN DCDC converter efficiency test
For this demo i made simple eGaN buck converter based on two 100V 48A 4mOhm EPC2032 eGaN transistor. They are very very very fast, can run up to 1-10MHz with correct driver, have no body diode, very low gate charge and so on. And gate voltage is only 5V (6V will kill it)
Note that for proper layout, i need microvias and other fancy pcb technology, only then you can get 95+ efficieny. My layout was best i can do with 4 layer oshpark service ( since pcb cost is at least 400eur even for mall ones, if you want correct layout)
On the right side you see PWM high and PWM Low signals that is generated by STM32F429.
In this test, i set input voltage of 10V, and limited current to 4A if something goes south. with STM32F4 i set around 3.3Vout, and measured efficiency by calculating power in and out. I will have many loses on cables, but that's ok. Here is what i get for 100kHz, 200kHz, and 1MHz. I did pushed PWM to 5MHz, but it was heating quite good at 2Ainput. This is only because of bad layout, eGaN can work at this frequency no problem with correct driver, i can't stress that enough.
So this test was so easy to do with this electronic load. It is such a joy to work with it. I could load my device to any current i desire, and it just keeps that current no matter what. Wery happy how this instrument is performing, and feedback it is giving.with power supply i had, it max out at 13A from 10V 5A power supply. Note that this power supply is very tiny, only 1.5cmx1.5cm and still can push some heavy current ! Not only that it does support voltages up to 100V 
By changing input voltage to 20V, was pushing 84W
In order to regain some eGaN trust i soldered my board that has more correct pcb. Driver is the same, but in BGA package, also layout is highly optimized. it use tones of 0.15/0.3mm vias to transfer power heat, and reduce inductance. And yes, it does work much better. From same size of 1.5cmx15cm i was pushing 120W 12V 10A from 25V 5A power supply, and it was in thermal equilibrium. Note that eGaN can work at 100C no problem, this is the one of advantages of eGaN devices, very high working temp.So after 10min of work, i produces this graph of dc/dc converter efficiency, it is maxing at 95-97% efficiency at very high loads. Note that pcb area is very small for this DCDC converter !!!!
Next test was piezo driver that i am working in my job. This is downsized demo unit just for lab use. It has slew rate limited 100V dc/dc converter that is running from 12-24V. Due to very low noise, it is designed not for efficiency. Slew rate limiting gives very little ringing or other type noise, because where are no high frequency content, but at the cost of heat in mosfet's. I had a suspicion it has some sweat spot where it is most efficient, and i was right, it is clearly visible from efficiency graph
So this test show how good my converter is, and i should use 13V power supply if i can, but due to restrict in my system, i have to use 24V.... Bummer
Battery testing
For this test from ebay i get the best battery ever, it is so light, and have 10Ah capacity. That would be very nice battery from quadcopters. Since we have perfect tool for the job, lets's test it ! And yes, no surprice it get 0.5Ah capacity at 1V cutoff voltage, that is just peace of ****.
Review still under construction
Top Comments
Thank you very much for this review, Linas!
I'm also building a custom PCB for GaN devices. I was getting a quote of approx. 70€ for a four layer board . And 100€ more if I would order the correct copper…