LR8 reg: a handy utility block

I’ve used the LR8 reg a few times here and there, always with satisfactory results. Yet, I’ve never spent time looking into it with more detail, always helped me to fix a voltage reference needed. Over time, I shifted towards the gas regulators and MOSFET-based voltage regulators where needed.

I also wrote extensively about the “sound” of regulators and why I’m reluctant (as many out there) to include closed loop regulators in my HT supplies. Having said that, having a tight closed loop reg to generate a reference and use a cap multiplier is a great alternative to provide a very clean supply.

The LR8 can do a great job here for voltages below 430V-ish if you marry him with a nice MOSFET like this:

LR8 regulator

The LR8 needs a 1.2V reference across the ADJ and output pins. R1 and Rtrim+R3 provide the adjusting network to dial the output voltage. See the datasheet for more info. R3 sets the minimum voltage and Rtrim provides the range needed for adjustment. Of course you can tweak this network for your needed voltage. I just simply worked this out to provide a 90V reference. RLPF and CLPF are the low pass network feeding the pass MOSFET. The LR8 has some BW noise like every reg which spreads into HF. Although there is no info on the datasheet, I remember measuring this but can’t find the data at hand. The 220K+47uF combination on the LPF network seem to provide a good balance on the rejection at 50Hz as well as the transition response to slowly rise the output in about 45sec.

The spice model I use isn’t trustworthy as is not official. It shows a notch which can be moved to about 100Hz with the above RC network values chosen. Proof is in the pudding and have to measure this to see if it’s somewhere near this simulation. Either way, the rejection is over 80dB across the band and if it measures as good as this, then we are all happy.

I’ll do a small PCB for this so if there are interested on it, just let me know.

Swenson+ tests with SMPS

Here is a test you might want to take into account. I own an HGC-320-700A SMPS which only allows for maximum current trimming, but output voltage is fixed. In order to avoid unnecessary heatsink, I will suggest using an RC before the Swenson where the R is a big Alu clamped wirewound resistor bolted to the chassis. Just leave 50V for the regulator which may dissipate a lot anyway depending on the output current.
 
There is no caps below and was stable enough. I could have added an output cap but I didn’t on this test. You can see how all the HF noise is trimmed out! About 40 to 60dB 🙂
 
The 44mA and 88mA traces are the SMPS feeding the power resistor directly. 
 
 
This can be improved further with a (mandatory) output cap. A nice 30uF oil cap should do well.
 
I have offered some pairs built and tested to builders using the prototype boards. I may do a couple more if you are interested