GWSGems #1 – Efficient output adjustment on a Clinac

In my time as a self-employed contractor I got to touch a lot of equipment and swap tips and clues with a lot of people. I have tended to keep many of the best of these to myself and my closest colleagues. Partly to protect valuable intellectual properties, partly to avoid putting potentially dangerous information in the hands of people not prepared to use it properly.

But anyone who has paid any attention also knows that I have a tendency to give away very valuable intellectual property out of a longstanding misguided idealistic impulse. I always hear the sage words of Fred Brooks on this point. When asked wasn’t he worried someone would steal his ideas he responded that it was his job to persuade people to steal his ideas. Idea propagation has to be its own reward since folks rarely recall where the idea came from.

So here’s #1 in a possible series. Today I will show you how to use the digital display of a Clinac to adjust output very quickly and efficiently. First the motions, then the theory.


Let’s say you have done your output check by whatever means and determined that your output is outside your acceptable tolerance. It should be Ointended (likely 1) but instead is Oactual. Let’s call the ratio R = (Oactual / Ointended).

You want your Clinac to be in service mode. I’m going to assume here that it is of a reasonably current vintage. I’ll say more about that at the end of this section.

  1. Turn on the digital parameter display (type DH). The offset is probably 67B.
  2. Run CCE.
  3. Note the dose rate displayed on the Clinac console. It will be something close to 2000. Let’s call it Dactual. Write it down or punch it into your calculator or spreadsheet.
  4. Calculate Dtarget = Dactual x R. Adjust the MU1 trim pot until the Clinac shows the dose rate to be equal to Dtarget.
  5. Adjust the ION1 trim pot until the Clinac shows the dose rate to be 2000.
  6. Stop CCE. Clear interlocks. Run some beam, maybe 2000 MU, enough to complete the next step.
  7. Looking at the digital parameter display adjust the MU2 trim pot until the second pair of hex values matches the first pair.
  8. Stop the beam. Start CCE.
  9. Looking at the digital parameter display adjust the ION2 trim pot until the second pair of hex values matches the first pair.
  10. Turn off CCE. Clear interlocks. The output is now adjusted.

You should run 200 MU to confirm that MU1 and MU2 displays track together and terminate at 200. You should also CCI and again check that MU1 and MU2 displays track together and terminate at 200. You should confirm that the output is now quite close to Ointended.

Some details

The digital parameter display shows current telemetry as hexadecimal values. I have not much idea what most of it is, but if you enter the proper starting offset then the first four 2-digit hex numbers are the low byte of dose rate channel 1, the high byte of channel 1, the low byte of channel 2 and the high byte of channel 2 respectively. When I talk about matching the values above, what you want to do is first make sure that the 4th number matches the second and then that the 3rd matches the first. Or something like that.

For older Clinac software versions some of the details are different. For instance, some older versions do not have CCE. There is a trick for setting the time of the CCI to be longer which is detailed in CTB-282 available from Varian.

For older versions of the integrator board, the dose rate magic number is 1050 rather than 2000. This is also discussed in CTB-282.

For older versions of the Clinac software the digital display offset to get to the dose rate values is different from 67B. Here is a table cribbed from a Varian training manual c. 2000.

hex offsets

The theory

I will share with you how I conceive what is going on. It’s a useful model but should not be mistaken for actually true in any particular detail.

So, conceptually, the integrator circuit consists of a signal source and an integrator with variable sensitivity. I like to think of the MU1 or MU2 trim pot as a gain control on the integrator input. What we have to do in order to control the actual dose per MU delivered at a reference point is to adjust the gain so that the desired number of MU are counted for a specific input signal. (I frankly don’t know how the integrator works but one way or another it has to produce a signal that is proportional to the charge collected in the monitor chamber cell.)

The Clinac also has a self-test mechanism for the integrator that involves a fixed reference signal. At mode-up (or at CCI or CCE) the reference signal is directed to the integrator in place of the monitor chamber signal. After the gain of the integrator input is set, the reference signal has to be adjusted to give the correct signal for the desired number of pseudo-MU, nominally 2000. I think of the ION1 and ION2 trim pots as controls on the value of the reference signal.

So… What we are doing in my procedure above is the following:

  1. Find the current value that results from integrating the current reference signal at the current integrator input gain.
  2. Adjust the integrator gain for channel 1 by the factor required to produce the desired MU/charge. Logically, if the output is currently too high, then the gain needs to be adjusted to produce more MU for a given input signal.
  3. Set the reference signal for channel 1 so that it produces the standard number of MU at the newly-adjusted integrator input gain.
  4. Use the digital display to set the channel 2 gain and the channel 2 reference value to match channel 1.

Note that order is of the essence. The reference signal must be set after the integrator gain.


Both Joe Presser and Ken Chu had key pieces of this puzzle and were gracious enough to share. My good friend Jim Gaiser served as guinea pig when I first tried to explain this to someone else.

This entry was posted in Wisdom and truth. Bookmark the permalink.

1 Response to GWSGems #1 – Efficient output adjustment on a Clinac

  1. flavio says:

    Hello, this is only to made public an e-mail sent to Varian on 02/28/2011, without any answer up to date, where I pointed out that this procedure was developed in an independent way by me and shared inocently to whom that wanted to heard about:

    Having working ever with digital machines I’ve got the habitude to perform simple procedures in order to change the gain of the MU counting device to have the desired MU/cGy. In preparation for the arrival of our new Clinacs I went to follow HE-Clinac-TM1 and 2 (January 2009) and there I was confronted to the fact that there was no a simple way to do MU/cGy calibration. Actually I asked explicitly about that and the only hint I received from trainers was to look after the hex dose rate.
    It was in Las Vegas where, thinking about this problem, I realized that one can use “Extended integration” to change the MU gain in a controlled and “human” way (seeing decimal dose rate indication instead of hexadecimal dose rate). Back to my hospital I implanted this technique from the very first use of our clinacs and I taught my physicist colleges in doing so. In December 2009 my college Maxime Villeneuve went to Las Vegas to follow TM1 course and there he was faced the same situation: people asking how to calibrate MU1 in an easy way and again the same hints: to watch the hex dose rate to do the job. There my college showed publicly our very simple way of doing such calibration.
    As far as I know no other hospital was using our method up to that date.
    After 3 moths that my college made public our method Varian approves the release of CTB CN-282 were the technique of using “extended integration” is made public.
    It could be possibility that Varian has developed such technique in an autonomous way but it is very suspecting that it has been made public only after my college publicized it during the TM1, mostly taking in account that in every Varian document prior to this one no any method was suggested to do that job in an easy way and that was always declared as being the sole responsibility of the customer.
    I am concerned about the fairness of this situation: I have the strong doubt that someone inside Varian has realized of the simplicity of that technique and about the importance of made it public but forgot to mention what the source was.
    If it is the case we want to be recognized for that, with the explicit mention that this technique was developed by me as part as my work at the Gatineau Hospital.
    Thank you very much for your attention.
    Best regards,

    Flavio Enrico Nelli
    Physicien M├ędical

Leave a Reply