• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Using SCR to maintain on condition, for a while.

Status
Not open for further replies.

kinarfi

Well-Known Member
Untitled.png
I've been trying to use and SCR to hold a circuit off after the current has reached a set level and cause a comparator to trip and fire the SCR, which should stay tripped, until turned off by another set of conditions. The problem I'm running into is that even when turned on, the SCR still has the customary .7 volt drop, which is enough other components in the circuit to turn on or not turn on as desired. Please look at the simplified circuit, is there another scheme for the use of SCRs? If I add voltage between the SCR and ground in the form of a diode, zener or LED to make the trigger voltage be higher, it also makes it easier for the input to turn it off , my best luck has been to put a LED in the gate lead to force the triggering voltage to above the off plus noise level of the triggering device which also has the .7 volt or more minimum.
It SIMs good, but so far I've had to rebuild the circuit several times and replace the U8 Power PFET a few times, even though the power supply is limited to about 7 amps.
The SCR is supposed to trip when the current from the motor reaches the end of travel and stalls causing the current to rise and stay tripped until the call for the steering motor is removed, causing the and gate to activate.
The above file is 000 Practice 16-12-02.asc, The entire schematic is 000 STEERING-RELAY 16-12-07 CORE + EOT.asc, but you probably won't be able to run the simulation due to components you won't have.
I will appreciate any suggestions or advice you may have, the SCR problem seems to be the last stumbling block, the rest seems to be working great.
Jeff
 

Attachments

Last edited:

kubeek

Well-Known Member
Most Helpful Member
Why do you use an SCR? Wouldn´t it be much easier to do all the function using a single S-R latch, and maybe a few other gates? I.e. a call sets the latch, and an overcurrent resets it.

If you could descibe in points what the circuit is supposed to do, including what that and gate is supposed to do and why is there an optocoupler, it would be a great starting point.
 

alec_t

Well-Known Member
Most Helpful Member
Agree with kubeek.
A FET would make a lower-voltage pull-down than an SCR.
 

AnalogKid

Well-Known Member
Most Helpful Member
The LEM sensors I'm familiar with source an output current, and you scale that to whatever volts/amp with a resistor to GND. Your circuit has a pull *up* resistor on the LEM output. What is the sensor part number?

And I agree with the others. For the lowest gate voltage, replace the SCR with a 2N4401, 2N7002, etc., and drive that with latch logic.

ak
 

ronsimpson

Well-Known Member
Most Helpful Member
M4 turns off at a very low voltage. The SCR +M2 can not pull that low.
Here, with a resistor divider, I moved the bottom end of M4 up a little. (R3, R4)
R3 could be a diode so you have 0.7V.
M7 will have a slightly smaller turn on voltage but I think that is OK.
upload_2016-12-8_7-30-36.png
 

kinarfi

Well-Known Member
This is the end of travel detector for a H Bridge driven steering motor, it consists of wheat stone type input and an oscillator input fed to a LM339 quad comparator, (two of the outputs are fed to the and gate), that outputs are fed to 2 lm556 timers used a FET drivers, (this Part Works quite nicely) the current is monitored by the modified LEM HX 50-P/SP2 which is compared to a set voltage and trips the(this Is where I'm having trouble) SCR to turn off the P FET and stop all current to the motor and is reset by the and gate & opticoupler and every thing is back work. The P FET is also turned off when the key is off, power is always applied to the P FET, not my design.
I'm not intimately familiar with the S-R latch, does it come in DIP8, space problems. the opticouple was a handy way to use the output of the and gate and may be rethought, the Lem is HX 50-P/SP2, but I took it apart and made 4 loops around the Hall device, effectively making it a "HX 12.5-P/SP2 which has a 50mv /amp output from a 2.500 reference.
With a 100Ω below M4, it worsens my problem by raising Vd of M4 when turned on when it's purpose is to lower Vg of M7 turning it on harder, that may have been part of my problem because I was getting about 3 volts on the drain when I wanted 0, I'll have to look into that today, that could be what killed my FETS. I'll check back tonight
Jeff
 

ronsimpson

Well-Known Member
Most Helpful Member
I must not understand the problem(s).
M4 is not surviving. Does it get hot?
You say the power supply is limited to 7A but the 0.1 ohm resistor pulls 100 to 140A. So what happens? Does the supply drop to 5 volt?
I can't read what P-mosfet you are using? I think your supply collapse to 5V. Maybe the gate must be pulled 5V below the supply for the MOSFET to be on. With a supply of only 5V it can not be turned on well so it will heat and burn out.
 

ronsimpson

Well-Known Member
Most Helpful Member
Don't really understand what you are building but:
It looks like a battery discharger.
Here is a circuit you might get some ideas from.
I could not find a TL431 so I used a LT1431. V2 makes it work more like a TL431.
R5&R1 will cause M7 to latch open if the battery gets below 9.7V (change R5 to get the voltage you want)
Q1 monitors M7. If the voltage across it gets too big (0.65V) it will cause M7 to latch open. Crude way to monitor current.
I have LOAD set to 0.15ohms. At 0.1 ohms the circuit over currents and latches open. Change the RDSon of M7 to set current limit.
upload_2016-12-8_21-11-9.png
I used the TL431 only. Did not use the op-amp.
Made latch via M1, M4, M7, Q1. Could have made a latch more simple but wanted to add current limit.
Spice file has an error but will run.
Removed battery and added a PULSE voltage source so you can see the voltage drop with time.
 

Attachments

kinarfi

Well-Known Member
Ron,
The P Fet is IRF4905 55V-74A-20mΩ, and I killed it, the M4 BS170, 60 v 500 ma has survived fine, the power supply is a Lambda 20 v, 8 amp with volt and amp limit, and you caused me to look at something that hit nerve & I lost my concentration for a while, The sim is for use in a vehicle and the .1 ohm is in place of motor that stalls at about 50 amps.
Yes the power supply does drop and that is probably why I killed the P Fet.
As I started working on the board today, a crack in it became a break, so I'm rebuilding it.
The problem, as I see it , is with the SCR having the cathode grounded, although via NFet BS170, it seemed the other devices were tripping it with their normal bias voltage and to combat that, I did what you did, but with a diode, which raised it normal bais voltage to the next device. I ended up using an LED in the gate so the normal bias voltage was well below what it took to turn on the device. Another item is the VGS(th) Gate Threshold Voltage VDS = VGS, ID = 1mA All 0.8 2.1 3 V for the BS170, the Vak of the MCR100 is almost there so I wanted some room for what ever.
Working on the new board, I'll let you know , I'm doing the end of travel part first and when I'm happy with that, I'll do the rest.
Jeff
 

kinarfi

Well-Known Member
Don't really understand what you are building but:
It's a end of travel detector and current shut down. As the motor reaches it's end of travel and stalls, the current goes way up, so I set the trip point at a point well below stall and turn off the PFET supplying current to driver, to reset the SCR, the driver must be returned to 0 to turn off the the M3 FET (you're right, my schematic is hard to read)
The LEM is a hall effect current sensor that's modified.
 

Attachments

kinarfi

Well-Known Member
During the rebuilding of my circuit board, making sure every thing was working as I went along, the SCR problem was still there, that is it would trip, but it was resetting itself, the problem was not enough current to hold the latch up, I only had 2ma , testing showed the 13 ma was right on the verge of maybe hold, maybe not, changed the 5k Ω resistor to 470Ω 1\2 watt. It's working reliably now.
Jeff
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top