Previous in Forum: \$75.00 + \$53.88 to fill up the tank! Next in Forum: Falling ball rocket launcher
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105

# Help with analog circuit design for servo valve current drive

03/11/2022 6:16 PM

First off, let me state that this is not homework. At least it's not homework in the sense that I am not being graded for this. It is homework in the sense that this is a problem that's been bugging me off and on for 30 years (mostly off as it was something I didn’t quite understand 30 years ago, but it was only a minor issue and we were able to work around it, however, as time went on and I would go to a customer’s site to deal with quasi related problems, I would sometimes return back to trying to understand how this circuit worked, or was supposed to work).

Secondly, I am a degreed mechanical engineer who has worked in the controls/motion field since 1990. I do understand basic circuit theory and can work with and troubleshoot general analog circuit card, I lack understanding of nuances in certain areas, which is what I believe is causing me to not understand this design.

The design is an analog circuit to drive an electro-hydraulic servo valve which does not, DOES NOT, have a feedback wire in it. Software is used to close the loop. The circuit card has the following states of operation:

1. computer running, no damping, no playback
2. computer running, no damping, no playback
3. computer running with playback
4. computer not running,

The main two of concern are states 1 and 4.

There are switches that connect certain electrical pathways based on the above states. Those switches will either switch in a signal (servo command, position feedback, velocity feedback, force stabilization) or switch in a diode. The output of all the switches get summed together to feed an operational amplifier circuit which drives the current amplifier feeding the servo valve coils.

Some areas of confusion for me are the following:

1. What is the purpose of the diodes?

2. How does operation change if the diodes are oriented in the opposite direction? I ask because the board design this circuit was based on is essentially identical with the exception of the diodes being oriented the opposite way.

State 1

This circuit has the following components:

Operational amplifier U14A (SE5532) with these signals connected to the

• non-inverting input:
• ground through R55 (1K ohm)
• inverting input:
• ground through R90 (2K ohm)
• servo command (Servo Cmd) signal goes through R8 ( 52.4k ohm)
• a stabilizing input (band pass of a load cell signal) through R91 (10.5k ohm)
• bias offset signal (Potentiometer R51 and fixed resistor R52)
• Diodes (CR8, CR7 and CR10), 1N4148, to ground
• Feedback from servo valve coil via R53 (15k ohm)
• Output:
• R56 (100 ohm)

Current amplifier U48 (LH0002) with these signals connected to the:

• Input
• R56 (100 ohm) from U14A
• Output
• Servo valve coils (in parallel)

The way this state is supposed to operate is the servo command provides the primary signal for moving the servo valve. However, in my humble opinion, the 3 diodes will keep the inverting input signal essentially grounded thus not allowing the system to operate. That being said, the system DOES operate, so what am I not understanding?

State 4

This circuit has the following components:

Operational amplifier U14A (SE5532) with these signals connected to the

• non-inverting input:
• ground through R55 (1K ohm)
• inverting input:
• position feedback signal which goes through R89 (52.3k)
• velocity feedback signal which goes through R87 (4.99k)
• bias offset signal (Potentiometer R51 and fixed resistor R52)
• Diodes (CR11, CR12, CR9 and CR10), 1N4148, to ground
• Feedback from servo valve coil via R53 (15k ohm)
• Output:
• R56 (100 ohm)

Current amplifier U48 (LH0002) with these signals connected to the:

• Input
• R56 (100 ohm) from U14A
• Output
• Servo valve coils (in parallel)

The way this state is supposed to operate is when the computer is off, the servo command is switched out and one of the diodes (CR9) is switched in, the resistor, R90, is switched out and diode CR11 is switched in, the stabilizing circuit is switched out and diode CR12 is switched in. Diodes CR7 and CR8 are switched out and the position and velocity feedback signals are switched in. This provides an analog damped position feedback look control the valve which will in turn hold position of the actuator the valve is mounted to.

Again, in my humble opinion, the diodes will keep the inverting input signal essentially grounded thus not allowing the system to operate. That being said, the system DOES operate, so what am I not understanding?

I apologize if I’ve not described the operation clearly. It’s clear in my head, so I may have been too brief in my elaboration. Also, I noticed the images I've included are not very clear. I will see if I can figure out what format to use to get a clearer image.

__________________
J B
Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Guru

Join Date: Mar 2007
Location: at the beach in Florida
Posts: 30950
#1

### Re: Help with analog circuit design for servo valve current drive

03/11/2022 7:16 PM

Well I think the diodes limit the current flow to one direction, and the resistors limit current flow to ground...So let me get this straight, the switch is continuing to work with the computer off...? ...have you checked the diodes?..., they do go bad...they should have continuity in only one direction...

__________________
Break a sweat everyday doing something you enjoy
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
#2

### Re: Help with analog circuit design for servo valve current drive

03/11/2022 8:18 PM

I agree that the diodes limit current flow to one direction. That’s what confuses me. The direction the diodes are pointing means (unless I’m mistaken) means current is blocked from flowing from ground to the summing node and would allow all current from the servo command to flow to ground (i.e. rendering the servo command useless).
I have not checked the diodes as I am helping to troubleshoot the board remotely. I am troubleshooting a different problem and believe the board is functioning properly. I am just baffled by how it even functions if it is manufactured per the schematic (which it should be as the board traces and wave soldering are done based on the computer generated schematic).

I am also wonder if the design is incorrect but still able to function well enough for our purpose but just at a limited performance capability. I say that because the circuit card’s predecessor design showed the diodes installed in the opposite orientation. One of them has to be incorrect.?!

__________________
J B
Guru

Join Date: Mar 2007
Location: at the beach in Florida
Posts: 30950
#3

### Re: Help with analog circuit design for servo valve current drive

03/11/2022 9:23 PM

Well you can always flip them around and check it....but I would contact the manufacturers rep and ask them these questions, they probably have dealt with this issue many times...

__________________
Break a sweat everyday doing something you enjoy
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
#4

### Re: Help with analog circuit design for servo valve current drive

03/11/2022 9:46 PM

The boards were manufactured in the mid 90s by the company I work for. However since then, we no longer do in-house board design so all of our electronics engineers have left. There is no one knowledgeable of that design is left at the company. We used to have an in-house test bench we could put the board on and run various functional tests on it....but that was disposed of when we stopped manufacturing our own circuit cards.

__________________
J B
Guru

Join Date: Mar 2007
Location: at the beach in Florida
Posts: 30950
#5

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 12:27 AM

Well I would go though the personnel files to see if I could locate one of those engineers that does have knowledge of those boards, and hire them temporarily as a consultant, either that or declare the boards obsolete and hire a control engineer to redesign what you need...

__________________
Break a sweat everyday doing something you enjoy
Guru

Join Date: Aug 2005
Posts: 5266
#6

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 6:39 AM

Virtual ground

The diodes are just to clamp the signal close to ground under extreme circumstances. Three in parallel seems very odd: This would be more sensible

__________________
We are alone in the universe, or, we are not. Either way it's incredible... Adapted from R. Buckminster Fuller/Arthur C. Clarke
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
#7

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 7:48 AM

If their purpose in this circuit is to clamp the signal close to ground, WHY would this be done during normal operation (state 1 in my OP)? And how would the servo command actually do anything if it’s clamped to ground?

__________________
J B
Guru

Join Date: Aug 2005
Posts: 5266
#8

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 8:12 AM

Did you read the Wiki article on virtual ground?

Essentially when an inverting Op-amp is working correctly

The negative input is held at the same voltage as the positive input.

When the positive input is grounded the negative input is called a virtual ground.

I don't know why the diodes are there, but I'm guessing they are to clamp spikes or oscillations or something.

__________________
We are alone in the universe, or, we are not. Either way it's incredible... Adapted from R. Buckminster Fuller/Arthur C. Clarke
2
Guru

Join Date: Dec 2008
Location: Long Island NY
Posts: 15107
#9

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 4:11 PM

This is a classic current summing amplifier circuit with an output current buffer when this operates linearly. The diode array tells me this circuit is expected to operate briefly non-linearly from time to time. During non-linear operation, the three diodes will be forward biased by up to 0.7 volts and share the current load. During linear operation, the diodes will effectively be out of the circuit due to a near-zero voltage differential between this virtual ground node and the ground. I suspect this circuit becomes non-linear because the speed of the operational amplifier and current buffer are considerably faster devices than the servo coils can be changed.

__________________
"Don't disturb my circles." translation of Archimedes last words
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
#10

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 7:08 PM

Redfred, that seems reasonable to me.

How would the circuit behavior change if the diodes were pointing the opposite direction (i.e. with the anode towards ground)?

I've always assumed the original design (which had the diode's anode connected to ground) to be correct, but maybe I'm wrong about that. That board design was only used on few programs by the time I was utilizing them. It's possible that the schematics for the older design was incorrect and never updated when engineers determined a problem with functionality. The newer board design may have incorporated the changes and corrected the diode orientation. I might be able to dig through some documentation and find a trail confirming my hypothesis.

__________________
J B
2
Guru

Join Date: Dec 2008
Location: Long Island NY
Posts: 15107
#11

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 11:00 PM

The diodes, in either orientation, only speed up how quickly the circuitry comes out of non-linear operation. A 0.7-volt difference between the inverting and non-inverting operational amplifier inputs a poor open-loop voltage gain of 100 means the amplifier wants to produce the unachievable 70-volt output level. The op-amp front-end transistors will come out of saturation faster with a smaller voltage difference at the inputs. A pair of opposite orientation diodes allows a quicker transition out of saturation in either maximum output voltage.

The single orientation diode might be because the op-amp supply power of only one voltage and ground. In more technical words, VCC=+V from a single battery while VEE=GND.

Many newer op-amp chips have such back-to-back diodes built into the chips making these circuit board diodes superfluous.

__________________
"Don't disturb my circles." translation of Archimedes last words
Power-User

Join Date: Feb 2020
Posts: 172
#12

### Re: Help with analog circuit design for servo valve current drive

03/12/2022 11:41 PM

Diodes isolate and separate the inputs from each other.

__________________
The proper hardware will eliminate all theory.
Guru

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
#13

### Re: Help with analog circuit design for servo valve current drive

03/14/2022 9:47 AM

Randall,

I had not read it, but did after you mentioned it. It's funny what one (meaning me) has forgotten. It was helpful in refreshing my memory regarding analyzing Op Amp circuits.

__________________
J B
Interested in this topic? By joining CR4 you can "subscribe" to