Menu
What's new
Find's Treasure Forums

Welcome to Find's Treasure Forums, Guests!

You are viewing this forums as a guest which limits you to read only status.

Only registered members may post stories, questions, classifieds, reply to other posts, contact other members using built in messaging and use many other features found on these forums.

Why not register and join us today? It's free! (We don't share your email addresses with anyone.) We keep email addresses of our users to protect them and others from bad people posting things they shouldn't.

Click here to register!

Need Support Help?

Cannot log in?, click here to have new password emailed to you

Changed email? Forgot to update your account with new email address? Need assistance with something else?, click here to go to Find's Support Form and fill out the form.

709 Bandwidth

A

Anonymous

Guest
I have just found a couple of data sheets on the 709 amongst the dust and cobwebs in the attic. One is by Fairchild and the other, Raytheon. The only ones I can get now are National Semiconductor and these are becoming difficult to obtain. According to the Fairchild data, the gain falls to 0db at about 10MHz, which is more reasonable than the higher figure I quoted earlier. I use a gain of 45db with the 10p and 3.3p compensation capacitors (no R). This gives a flat response from dc to 2MHz which is adequate for the sampling delays encountered in TH detectors. Most times I now use the NE5534 which doesn't require external compensation, but is a bit slower for the same gain.
Eric.
 
Eric,
Have just sighted this minute. To be frank it was your comment re 500mhtz that prompted my bulldust entry to the forum.
Walter Jung in IC Opamp Cookbook,1980, page 83, shows frequency response for the 709 at various gains. (Copies Fairchild data)
I have been interested with your comment upon some modern chips being disappointing. Is this an area you can expand upon?
g
 
Bummer!!
My son, computer engineer, has happily just pointed out the error in my posted E-mail address. Hope this post will correct that. Todays play, substituting a larger and lower R mosfet very successful. (From IRF830 to IRFPC50) I'm guessing, but I think Eric uses the IRF740 which provides 0.55ohm. Todays challenge is to think through why my Tx damping values were not optimum with the mosfet change. Previously they were 3 x 620ohm 1w metal film. With the change of mosfet I took one R away - up to 310 ohms. The pulse looks so much quicker. Something going on here - avalanche voltage? Capacitance?
g.
 
What!!!
You do not like the topic???? Gotta find it before you hold it!!!!! Tell me what You do - tell me how - show me why - show me how you do it!!!!
g
 
L/R of the transmit coil and damping resistor is what is going on. Keeping uping the resistance till ringing and then back down a bit. Thats all.
JC
 
Hi JC,
Is it that simple? The characteristics of different mosfets having nothing to do with it? Eric mentions a three-step process, the first involving avalanche voltage. I suspect that the output capacitance of the mosfet is in there too - hence my earlier posting querying whether this was the case, and whether a diode would divorce this C from the inductance.
I may have misdirected myself, but I believe when changing the mosfet, that the optimum damping R value changed. L didn't change, and that only leaves C. That is where I am going today.
g
 
It is really that simple. Of course the coil L the mosfet C and other C are in parallel with the damping resistor R. Forms a bit of a complex network, and you can calculate the correct R, but is easier (and in the end you do this anyway) to raise the R until the return begins to ring (damped sinusodal) and then lower until it just quits, and a hair more, to make sure.
Try it, it will work.
JC
 
Actually Graeme, I was I that referred to the 709 saying it has (I think) in excess of 500MHz Gain-bandwidth (product). At the gain of 1000 used in the XL 500 this yields >500MHz/1000 = >500KHz bandwidth. This is still a very good number and is difficult to achieve with most other op-amps. The falling edge after diode clipping is very fast and is slowed considerably by op amps like the 5534.
 
Hi Charlie,
Thankyou. The discussion re 709 initially floored me and I posted before I located my copy of Walt Jung's book - which I recalled detailed the device extensively.
I stand corrected.
My purchase of 709's arrived yesterday. You may have seen I have raised a query - I would like to know just what defines a good or mediocre op-amp.
g
 
Hi Charlie,
I saw your post and thought the figure of 500MHz was a bit high and looked up a Raytheon application note on the 709. It gave a graph of open loop frequency responses for different values of compensation which stopped at 1MHz. However, if you extrapolated by continuing the ruler straight falloff and extending the base line it crossed 0db at 500MHz. However, I made the point that there would probably be other factors which would roll it off quicker. I use a 1M0 feedback resistor and the fraction of a pf that this has would add to the roll off, plus pcb capacitances etc. When I subsequently found the Fairchild data sheet the picture was rather different. Gone was the straight slide towards 0db and in its place was a curve whose slope gradually increased after 1MHz. The Fairchild curve stops at 5MHz but if you extrapolate that and gently increase the slope, I reckon that you would hit 0db between 10 and 20MHz.
For those that think I am living in the past as regards front ends, the one that I like for really fast work is the OPA620. This has a gain-bandwidth of 200MHz, is internally compensated and exhibits no peculiarities when overdriven, as some other high spec op-amps do. The downside is that you can
 
Hi Graeme,
Do you find 709's easy to get? Who makes the ones you have acquired? When more companies were making them i.e. Texas, Motorola, Fairchild, National and several others, I found significant differences in performance. Mainly in overshoot and noise. The ones I get now are National and these seem consistently good. I believe that Piotr mentioned on a while back that they are still made in Poland.
Eric.
 
I hate to beat the horse to death but the Fairchild data sheet for the 709 in the back of Jungs IC OP AMP COOKBOOK does show a closed loop bandwidth of about 800KHZ when the 709 is set to a gain of 60 dB and compensated accordingly. This gives a GBW product of 800MHz which is about as good as it comes unless you resort to exotic amps with other bad characteristics. I don't know if you really need this much bandwidth or not.Eric, do you have any experience with limited bandwidth in this first stage?
 
Hi Charlie,
My Fairchild curve gives 500kHz at 60db gain, but lets not split hertz. I think the point is that the slope of the curve ever increases after this and it becomes invalid to just multiply the gain by the bandwidth. I find that 500kHz is not enough to be able to sample at 15uS and have to reduce the gain to 45db to get the bandwidth flat to about 2MHz. Many other factor affect this but for my coils, transmitters, damping etc., this is generally the case.
Eric.
 
Thanks Eric. Now I have a number to work for. The XL500 is not settled much before 22-25uS and the 709 may be part of that. I suspect all the other factors are working against me here as well and I probably should just start from scratch.How much of an improvement should I expect on fine gold when going from say 25uS to 15uS- if all other factors were equal?
 
Hi Eric,
My purchase was via Farnell - fabulous and quick service, and no minimum order numbers. Yes, they are National - T05 case - I plumbed forLM709CH - at under half the cost of the LM709AH.
g.
 
Eric,
I do not accept this argument. You say balanced amplifier input only divorces noise upon thecommon-mode input? i.e. the coupling cable. Isn't that a valid argument initself todoso???? Please don't tell me RFI chooses to infect a damped coil or a nice wiggly ()just why do they do that!!!!} lead. Noise is noise. Are endeavours to suspret it so wrong???
g.
 
You must log in or register to reply here.
Top